graphical representation of air pollution

Air Pollution

Our overview of indoor and outdoor air pollution.

By: Hannah Ritchie and Max Roser

This article was first published in October 2017 and last revised in February 2024.

Air pollution is one of the world's largest health and environmental problems. It develops in two contexts: indoor (household) air pollution and outdoor air pollution.

In this topic page, we look at the aggregate picture of air pollution – both indoor and outdoor. We also have dedicated topic pages that look in more depth at these subjects:

Indoor Air Pollution

Look in detail at the data and research on the health impacts of Indoor Air Pollution, attributed deaths, and its causes across the world

Outdoor Air Pollution

Look in detail at the data and research on exposure to Outdoor Air Pollution, its health impacts, and attributed deaths across the world

Look in detail at the data and research on energy consumption, its impacts around the world today, and how this has changed over time

See all interactive charts on Air Pollution ↓

Other research and writing on air pollution on Our World in Data:

Air pollution: does it get worse before it gets better?
Data Review: How many people die from air pollution?
Energy poverty and indoor air pollution: a problem as old as humanity that we can end within our lifetime
How many people do not have access to clean fuels for cooking?
What are the safest and cleanest sources of energy?
What the history of London’s air pollution can tell us about the future of today’s growing megacities
When will countries phase out coal power?

Air pollution is one of the world's leading risk factors for death

Air pollution is responsible for millions of deaths each year.

Air pollution – the combination of outdoor and indoor particulate matter and ozone – is a risk factor for many of the leading causes of death, including heart disease, stroke, lower respiratory infections, lung cancer, diabetes, and chronic obstructive pulmonary disease (COPD).

The Institute for Health Metrics and Evaluation (IHME), in its Global Burden of Disease study, provides estimates of the number of deaths attributed to the range of risk factors for disease. 1

In the visualization, we see the number of deaths per year attributed to each risk factor. This chart shows the global total but can be explored for any country or region using the "change country" toggle.

Air pollution is one of the leading risk factors for death. In low-income countries, it is often very near the top of the list (or is the leading risk factor).

Air pollution contributes to one in ten deaths globally

In recent years, air pollution has contributed to one in ten deaths globally. 2

In the map shown here, we see the share of deaths attributed to air pollution across the world.

Air pollution is one of the leading risk factors for disease burden

Air pollution is one of the leading risk factors for death. But its impacts go even further; it is also one of the main contributors to the global disease burden.

Global disease burden takes into account not only years of life lost to early death but also the number of years lived in poor health.

In the visualization, we see risk factors ranked in order of DALYs – disability-adjusted life years – the metric used to assess disease burden. Again, air pollution is near the top of the list, making it one of the leading risk factors for poor health across the world.

Air pollution not only takes years from people's lives but also has a large effect on the quality of life while they're still living.

Who is most affected by air pollution?

Death rates from air pollution are highest in low-to-middle-income countries.

Air pollution is a health and environmental issue across all countries of the world but with large differences in severity.

In the interactive map, we show death rates from air pollution across the world, measured as the number of deaths per 100,000 people in a given country or region.

The burden of air pollution tends to be greater across both low and middle-income countries for two reasons: indoor pollution rates tend to be high in low-income countries due to a reliance on solid fuels for cooking, and outdoor air pollution tends to increase as countries industrialize and shift from low to middle incomes.

A map of the number of deaths from air pollution by country can be found here .

How are death rates from air pollution changing?

Death rates from air pollution are falling – mainly due to improvements in indoor pollution.

In the visualization, we show global death rates from air pollution over time – shown as the total air pollution – in addition to the individual contributions from outdoor and indoor pollution.

Globally, we see that in recent decades, the death rates from total air pollution have declined: since 1990, death rates have nearly halved. But, as we see from the breakdown, this decline has been primarily driven by improvements in indoor air pollution.

Death rates from indoor air pollution have seen an impressive decline, while improvements in outdoor pollution have been much more modest.

You can explore this data for any country or region using the "change country" toggle on the interactive chart.

Interactive charts on air pollution

Murray, C. J., Aravkin, A. Y., Zheng, P., Abbafati, C., Abbas, K. M., Abbasi-Kangevari, M., ... & Borzouei, S. (2020). Global burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019 . The Lancet , 396 (10258), 1223-1249.

Here, we use the term 'contributes,' meaning it was one of the attributed risk factors for a given disease or cause of death. There can be multiple risk factors for a given disease that can amplify one another. This means that in some cases, air pollution was not the only risk factor but one of several.

Cite this work

Our articles and data visualizations rely on work from many different people and organizations. When citing this topic page, please also cite the underlying data sources. This topic page can be cited as:

BibTeX citation

Reuse this work freely

All visualizations, data, and code produced by Our World in Data are completely open access under the Creative Commons BY license . You have the permission to use, distribute, and reproduce these in any medium, provided the source and authors are credited.

The data produced by third parties and made available by Our World in Data is subject to the license terms from the original third-party authors. We will always indicate the original source of the data in our documentation, so you should always check the license of any such third-party data before use and redistribution.

All of our charts can be embedded in any site.

Our World in Data is free and accessible for everyone.

Help us do this work by making a donation.

OUR NATION'S AIR

Trends through 2020.

For more than 50 years, the U.S. Environmental Protection Agency (EPA) has maintained its commitment to protecting public health by reducing pollutant emissions and improving air quality. This annual report, titled Our Nation's Air , summarizes the nation's air quality status and trends through 2020.

Sections of this report convey information across different time periods, depending on the underlying data sources. While some are consistently available since 1970, like growth data, our longer-term trends for air quality concentrations start in 1990, when monitoring methodologies became more consistent.

Please read and enjoy the full report below, and be sure to download and share the one page summary using the share button at the top. Additional detail on air trends can be found at EPA's AirTrends website.

Scroll down to read more or use the top menu to jump to a topic. If you encounter any issues viewing content, update or try opening the website in another browser.

United States Environmental Protection Agency

Since 1970, implementation of the Clean Air Act and technological advances from American innovators have dramatically improved air quality in the U.S. Since that time, the combined emissions of criteria and precursor pollutants have dropped by 78%. Cleaner air provides important public health benefits, and we commend our state, local, and industry partners for helping further long-term improvement in our air quality.

Air Quality Trends Show Clean Air Progress

Nationally, concentrations of air pollutants have dropped significantly since 1990:

Carbon Monoxide (CO) 8-Hour, 73%
Lead (Pb) 3-Month Average, 86% (from 2010)
Nitrogen Dioxide (NO 2 ) Annual, 61%
Nitrogen Dioxide (NO 2 ) 1-Hour, 54%
Ozone (O 3 ) 8-Hour, 25%
Particulate Matter 10 microns (PM 10 ) 24-Hour, 26%
Particulate Matter 2.5 microns (PM 2.5 ) Annual, 41% (from 2000)
Particulate Matter 2.5 microns (PM 2.5 ) 24-Hour, 30% (from 2000)
Sulfur Dioxide (SO 2 ) 1-Hour, 91%
Numerous air toxics have declined with percentages varying by pollutant

Despite increases in air concentrations of pollutants associated with fires, carbon monoxide and particle pollution, national average air quality concentrations remain below the current, national standards.

Air quality concentrations can vary year to year, influenced not only by pollution emissions but also by natural events, such as dust storms and wildfires , and variations in weather.

Emissions of key air pollutants continue to decline from 1990 levels:

Carbon Monoxide (CO), 70%
Ammonia (NH 3 ), 8%
Nitrogen Oxides (NO x ), 68%
Direct Particulate Matter 2.5 microns (PM 2.5 ), 38%
Direct Particulate Matter 10 microns (PM 10 ), 31%
Sulfur Dioxide (SO 2 ), 92%
Volatile Organic Compounds (VOC), 48%

In addition, from 1990 to 2017 emissions of air toxics declined by 74 percent, largely driven by federal and state implementation of stationary and mobile source regulations, and technological advancements from American innovators.

Wildfire data excluded for all pollutants except for NH 3 pre-2002; PM emissions also exclude miscellaneous emissions (i.e., agricultural dust and prescribed fire data). Visit the emissions trends website to learn more.

Tip Click pollutant names in the chart legend to hide or include trend lines, and hover over any line to display percentages above or below the most recent standard. Click the Emission Totals tab to view emission trends.

Air Pollution Includes Gases and Particles

Air pollution consists of gas and particle contaminants that are present in the atmosphere. Gaseous pollutants include sulfur dioxide (SO 2 ), oxides of nitrogen (NO x ), ozone (O 3 ), carbon monoxide (CO), volatile organic compounds (VOCs), and certain toxic air pollutants. Particle pollution (PM 2.5 and PM 10 ) includes a mixture of compounds that can be grouped into five major categories: sulfate, nitrate, elemental (black) carbon, organic carbon and crustal material.

Some pollutants are released directly into the atmosphere while other pollutants are formed in the air from chemical reactions. Ground-level ozone forms when emissions of NO x and VOCs react in the presence of sunlight. Air pollution impacts human health and the environment through a variety of pathways.

Six Common Pollutants

The Clean Air Act requires EPA to set national ambient air quality standards (NAAQS) for specific pollutants to safeguard human health and the environment. These standards define the levels of air quality that EPA determines are necessary to protect against the adverse impacts of air pollution based on scientific evidence. EPA has established standards for six common air pollutants, which are referred to as “criteria” pollutants.

Carbon monoxide (CO)
Nitrogen dioxide (NO 2 )
Ozone (O 3 )
Particulate matter (PM), and
Sulfur dioxide (SO 2 )

Understanding Emission Sources Helps Control Air Pollution

Generally, emissions of air pollution come from

stationary fuel combustion sources (such as electric utilities and industrial boilers),
industrial and other processes (such as metal smelters, petroleum refineries, cement kilns and dry cleaners),
highway vehicles, and
non-road mobile sources (such as recreational and construction equipment, marine vessels, aircraft and locomotives).

As the chart shows, pollutants are emitted by a variety of sources. For example, electric utilities, part of the stationary fuel combustion category, release SO 2 , NO x and particles.

Tip Click source categories in the chart legend to hide or include, and hover over any bar to display totals or percentages by source category. Click the ellipsis in the upper righthand corner and check "Show Totals" to view the chart based on totals instead of percentages.

Year:

Emission Inventories

EPA and states track direct emissions of air pollutants and precursor emissions, which are emissions that contribute to the formation of other pollutants in the atmosphere. Emissions data are compiled from many different organizations, including industry and state, tribal and local agencies. Some emissions data are based on actual measurements while others are estimates. For more information, please visit the Air Emissions Inventories website.

Air Pollution Can Affect Our Health and Environment in Many Ways

Numerous scientific studies have linked air pollution and specific pollutants to a variety of health problems and environmental impacts. Depending on the pollutant, people at greater risk for experiencing air pollution-related health effects may include older adults, children and those with heart and respiratory diseases — 30-second Healthy Heart video.

Health Effects Breathing elevated levels of CO reduces the amount of oxygen reaching the body’s organs and tissues. For those with heart disease, this can result in chest pain and other symptoms leading to hospital admissions and emergency department visits.

Environmental Effects Emissions of CO contribute to the formation of CO 2 and ozone, greenhouse gases that warm the atmosphere.

Health Effects Air toxics may cause a broad range of health effects depending on the specific pollutant, the amount of exposure, and how people are exposed. People who inhale high levels of certain air toxics may experience eye, nose and throat irritation, and difficulty breathing. Long term exposure to certain air toxics can cause cancer and long-term damage to the immune, neurological, reproductive, and respiratory systems. Some air toxics contribute to ozone and particle pollution with associated health effects.

Environmental Effects Some toxic air pollutants accumulate in the food chain after depositing to soils and surface waters. Wildlife and livestock may also be harmed with sufficient exposure. Some toxic air pollutants contribute to ozone and particle pollution with associated environmental and climate effects.

Health Effects Depending on the level of exposure, lead may harm the developing nervous system of children, resulting in lower IQs, learning deficits and behavioral problems. Longer-term exposure to higher levels of lead may contribute to cardiovascular effects, such as high blood pressure and heart disease in adults.

Environmental Effects Elevated amounts of lead accumulated in soils and fresh water bodies can result in decreased growth and reproductive rates in plants and animals.

Health Effects Oxides of nitrogen are a group of highly reactive gases, for which nitrogen dioxide (NO 2 ) is the gas of greatest health concern. Short-term exposures to NO 2 can aggravate respiratory diseases, particularly asthma, leading to respiratory symptoms, hospital admissions and emergency department visits. Long-term exposures to NO 2 may contribute to asthma development and potentially increase susceptibility to respiratory infections.

Environmental Effects Oxides of nitrogen react with volatile organic compounds to form ozone and react with ammonia and other compounds to form particle pollution resulting in associated environmental effects. Deposition of oxides of nitrogen contribute to the acidification and nutrient enrichment (eutrophication, nitrogen saturation) of soils and surface waters, ozone formation, as well to the direct and indirect effects on vegetation, soils, and animals.

Health Effects Ozone exposure reduces lung function and causes respiratory symptoms, such as coughing and shortness of breath. Ozone exposure also aggravates asthma and lung diseases such as emphysema leading to increased medication use, hospital admissions, and emergency department visits. Exposure to ozone may also increase the risk of premature mortality from respiratory causes. Short-term exposure to ozone is also associated with increased total non-accidental mortality, which includes deaths from respiratory causes.

Environmental Effects Ozone damages vegetation by injuring leaves, reducing photosynthesis, impairing reproduction and growth and decreasing crop yields. Ozone damage to plants may alter ecosystem structure, reduce biodiversity and decrease plant uptake of CO 2 . Ozone is also a greenhouse gas that contributes to the warming of the atmosphere.

Health Effects Exposures to PM, particularly fine particles referred to as PM 2.5 , can cause harmful effects on the cardiovascular system including heart attacks and strokes. These effects can result in emergency department visits, hospitalizations and, in some cases, premature death. PM exposures are also linked to harmful respiratory effects, including asthma attacks.

Environmental Effects Fine particles (PM 2.5 ) are the main cause of reduced visibility (haze) in parts of the U.S., including many national parks and wilderness areas. PM can also be carried over long distances by wind and settle on soils or surface waters. The effects of settling include: making lakes and streams acidic; changing the nutrient balance in coastal waters and large river basins; depleting the nutrients in soil; damaging sensitive forests and farm crops; and affecting the diversity of ecosystems. PM can stain and damage stone and other materials, including culturally important objects such as statues and monuments.

Health Effects Among the species of SO x , SO 2 is the most commonly occurring in the atmosphere and the one most clearly associated with human health effects. Short-term exposures to SO 2 are linked with respiratory effects including difficulty breathing and increased asthma symptoms. These effects are particularly problematic for asthmatics while breathing deeply such as when exercising or playing. Short-term exposures to SO 2 have also been connected to increased emergency department visits and hospital admissions for respiratory illnesses, particularly for at-risk populations including children, older adults and those with asthma. SO 2 contributes to particle formation with associated health effects.

Environmental Effects Sulfur oxides react with ammonia and other compounds to form particle pollution resulting in associated environmental effects. Deposition of sulfur oxides contributes to the acidification of soils and surface waters and mercury methylation in wetland areas. At certain concentrations, sulfur oxides can also cause injury to vegetation and species loss in aquatic and terrestrial systems.

For over 50 years, the Clean Air Act has played a major role in cutting pollution as the U.S. economy has grown. Despite the sharp impacts from the COVID-19 pandemic on activity in 2020, the U.S. economy remained strong.

Economic Strength with Cleaner Air

Between 1970 and 2020, the combined emissions of the six common pollutants (PM 2.5 and PM 10 , SO 2 , NO x , VOCs, CO and Pb) dropped by 78 percent. This progress occurred while U.S. economic indicators remain strong.

To learn more about the EPA and environmental milestones to reduce pollution please visit the EPA history website.

Tip Click any of the legend items on the right side of the chart to hide or include trend lines. The y-axis may change based on the selections.

In 2008, the United States environmental technologies and services industry supported 1.7 million jobs. The industry generated approximately $300 billion in revenues and exported goods and services worth $44 billion - larger than exports of sectors such as plastics and rubber products. Environmental technology exports help the U.S. balance of trade, generating a $10.9 billion surplus in 2008.

National Ambient Air Quality Standards (NAAQS)

For more than 50 years, the Clean Air Act has brought Americans cleaner air and lower risks of adverse health effects.

Criteria Pollutant Trends Show Clean Air Progress

Year:
# of National Stats Sites:
Maximum:
90th Percentile:
10th Percentile:
Minimum:
National Average:
Total:

Site Name:
Location:

Unhealthy Air Days Show Long-Term Improvement

The Air Quality Index (AQI) is a color-coded index EPA uses to communicate daily air pollution for ozone, particle pollution, NO 2 , CO and SO 2 . A value in the unhealthy range, above the national air quality standard for any pollutant, is of concern first for sensitive groups, then for everyone as the AQI value increases. Fewer unhealthy air quality days means better health, longevity, and quality of life for all of us.

Tip Shown are the number of days in which the combined ozone and PM 2.5 AQI was unhealthy for sensitive groups (orange) or above (red, purple or maroon) for the years 2000-2020. Click the bar chart, or these links, to view AQI retrospective reviews: PM 2.5 or ozone.

Unhealthy air quality days vary year to year, influenced not only by pollution emissions but also by natural events, such as dust storms and wildfires , and variations in weather.

A look back: Combined Ozone and PM 2.5 in 2020

Air Quality Index (AQI) Forecast

EPA provides a daily AQI forecast so people can act to protect their health. Shown is the current AQI forecast for PM and ozone combined. This map and others can be found at the AirNow website.

Air Quality in Nonattainment Areas Improves

EPA works collaboratively with state, local and tribal agencies to identify areas of the U.S. that do not meet the national ambient air quality standards (NAAQS). These areas, known as nonattainment areas, must develop plans to reduce air pollution and attain the NAAQS.

Through successful state led implementation, numerous areas across the country are showing improvement and fewer areas are in nonattainment. Since 2010, there were no violations of the standards for NO 2 .

Tip Shown is a snapshot of the 2008 ozone nonattainment area map. Click the map to view a larger interactive version that includes all current NAAQS nonattainment areas.

Nonattainment Areas

Over its 50+ year history, EPA has made significant progress in protecting the magnificent views of America’s national treasures from pollution. State and federal governments are working together to improve the natural visibility in our nation’s parks and wilderness areas so that future generations can enjoy these scenic vistas.

Visibility Improves in Scenic Areas

EPA and other agencies, such as the National Park Service, monitor visibility trends in 155 of the 156 national parks and wilderness areas (i.e., Class I areas).

The map indicates most Class I areas have improving visibility or decreasing haze (indicated by the downward pointing arrows). To learn more about visibility in parks and view live webcams please visit this National Park Service website.

Tip Click any point to display 2000-2019 trends, and select maximize to enlarge the chart. Double click the map to zoom in and click the home button to reset.

Site ID: Site Name: National Stats Site: Overall Trend:

Regional Haze Rule

The Regional Haze Rule, published in 1999, requires states to identify the most effective means of preserving conditions in Class I areas when visibility is at its best (based on the 20% best or clearest visibility days monitored) and to gradually improve visibility when it is most impaired (based on the 20% worst visibility days monitored).

Following the 1990 Clean Air Act Amendments, significant improvements in public health protection occurred as a result of reductions in air toxics emissions from large industrial facilities and transportation.

Air Toxics Levels Trending Down

Ambient monitoring data show that some of the toxic air pollutants, such as benzene, 1,3-butadiene and several metals, are declining at most sites.

Points on the map indicate the long-term statistical trend direction: decreasing, increasing and no trend. Depicted in gray are sites where a trend direction is undetermined due to insufficient data.

Tip Use the dropdown menu to select a pollutant, click any point to display trends, and select maximize to enlarge the chart. Double click the map to zoom in and click the home button to reset. View a tabular summary of air toxics trends.

Based on the 2014 NATA, secondary pollution formation is the largest contributor to cancer risks nationwide, accounting for 47 percent of the risk. On-road mobile sources contribute the most risk from directly emitted pollutants (about 12 percent).

Site Name: Location: NATTS: National Stats Site: Overall Trend:

Year:
# of National Stats Sites:
Maximum:
90th Percentile:
10th Percentile:
Minimum:
National Median:

Charts The play/pause button controls animation, or manually change the year by dragging the yellow circle in the chart or the slider's gray square. Read about weather influences on ozone. Few lead sites met trend completeness criteria to calculate national stats prior to 2010, and emissions data are only available for National Emissions Inventory (NEI) years.

Day:
Site ID:

Map Click any point to display daily concentration data. Double click the map to zoom in and click the home button to reset. Please be patient with map exports.

More Ways to Explore 2020 Air Quality

Check out the Daily Air Quality Tracker and Multiyear Tile Plot on the EPA’s AirData website for 2020 daily concentrations of criteria pollutants compared with historical values.

Air Quality Impacts From Wildfires

Map Symbols indicate values above or below the most recent standard. Click any point to display daily concentration data. Double click the map to zoom in and click the home button to reset. Please be patient with map exports.

Fire and Smoke Map

Explore the Fire and Smoke Map on the EPA’s AirNow website for up-to-date PM 2.5 air quality (including any wildfire impacts) as measured by a combination of AirNow monitors, U.S. Forest Service temporary monitors, and low-cost sensors.

Our Nation's Air Continues to Improve

However, work must continue to ensure healthy air for all communities. EPA and our partners at the state, tribal and local levels will continue to work to address the complex air quality problems we face.

Download and share the one page summary and scroll down for additional resources.

Social Media

Use the share button at the top to share this report with others, and follow the latest EPA activities to protect human health and the environment using the buttons below.

Source code, data and documentation are available for download in the GitHub repository.

Additional Resources

Please visit other EPA air quality related websites.

Air Emissions
>.export.print(' >',1500)" onkeydown="if (event.keyCode === 27) { >.export.hideMenu(); >.export.focusMenu()};if (event.keyCode === 13) >.export.print(' >',1500)" class="exportPrint"> Print
>.export.exportCSV(' >.csv')" onkeydown="if (event.keyCode === 27) { >.export.hideMenu(); >.export.focusMenu()};if (event.keyCode === 13) >.export.exportCSV(' >.csv')" class="exportCSV"> Export Data
>.export.exportPNG(' >.png',1500)" onkeydown="if (event.keyCode === 27) { >.export.hideMenu(); >.export.focusMenu()};if (event.keyCode === 13) >.export.exportPNG(' >.png',1500)" class="exportPNG"> Export PNG

Contributors
News and Events

State of Global Air Report 2024

The State of Global Air 2024 reports provides a comprehensive analysis of data for air quality and health impacts for countries around the world. The analysis finds that:

● Air pollution accounted for 8.1 million deaths globally in 2021, becoming the second leading risk factor for death, including for children under five years. Of the total deaths, noncommunicable diseases including heart disease, stroke, diabetes, lung cancer, and chronic obstructive pulmonary disease (COPD) account for nearly 90% of the disease burden from air pollution.

● In 2021, more than 700,000 deaths in children under 5 years were linked to air pollution; this represents 15% of all global deaths in children under five. As in previous years, the State of Global Air 2024 report and accompanying website provides comprehensive data on the levels and trends in air quality and health for every country in the world. This State of Global Air report was produced in partnership with UNICEF. In the State of Global Air 2024 interactive app , you can explore, compare, and download data and graphics reflecting the latest air pollution levels and associated burden of disease for over 200 individual countries, territories, and regions, as well as track trends from 1990 to 2021.

Fact sheets
Facts in pictures

Publications

Questions and answers
Tools and toolkits
Endometriosis
Excessive heat
Mental disorders
Polycystic ovary syndrome
All countries
Eastern Mediterranean
South-East Asia
Western Pacific
Data by country
Country presence
Country strengthening
Country cooperation strategies
News releases
Feature stories
Press conferences
Commentaries
Photo library
Afghanistan
Cholera
Coronavirus disease (COVID-19)
Greater Horn of Africa
Israel and occupied Palestinian territory
Disease Outbreak News
Situation reports
Weekly Epidemiological Record
Surveillance
Health emergency appeal
International Health Regulations
Independent Oversight and Advisory Committee
Classifications
Data collections
Global Health Observatory
Global Health Estimates
Mortality Database
Sustainable Development Goals
Health Inequality Monitor
Global Progress
World Health Statistics
Partnerships
Committees and advisory groups
Collaborating centres
Technical teams
Organizational structure
Initiatives
General Programme of Work
WHO Academy
Investment in WHO
WHO Foundation
External audit
Financial statements
Internal audit and investigations
Programme Budget
Results reports
Governing bodies
World Health Assembly
Executive Board
Member States Portal

The Global Health Observatory

Explore a world of health data

GHO OData API
Map Gallery
Data Search

Air pollution data portal

Air pollution data portal, air pollution data.

Burden of disease from air pollution

Ambient air pollution exposure

Household air pollution exposure

Interactive tools and resources

Visualization of national air quality standards

Clean Household Energy Solutions Toolkit (CHEST)

Household Multiple Emission Sources Model (HOMES)

Benefits of action to reduce household air pollution (BAR-HAP) tool

Performance Target (PT) Model

WHO global air quality guidelines: particulate matter (‎PM2.5 and PM10)‎, ozone, nitrogen dioxide, sulfur...

Clean air is fundamental to health. Compared to 15 years ago, when the previous edition of these guidelines was published, there is now a much stronger...

Personal interventions and risk communication on Air Pollution

This report presents the results of an expert consultation whose objectives were to review and assess the current scientific evidence on questions related...

Ambient air pollution: A global assessment of exposure and burden of disease

This report presents a summary of methods and results of the latest WHO global assessment of ambient air pollution exposure and the resulting burden of...

WHO Guidelines for indoor air quality: Household fuel combustion

In 2014, WHO issued the first-ever health-based guidelines on clean fuels and technologies for household cooking, heating and lighting. These guidelines...

Related teams

Feedback for the new GHO portal

The e-mail format is "[email protected]"

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

View all journals
Explore content
About the journal
Publish with us
Sign up for alerts
Open access
Published: 18 November 2021

Health impacts of air pollution exposure from 1990 to 2019 in 43 European countries

Alen Juginović 1 ,
Miro Vuković 2 ,
Ivan Aranza 2 &
Valentina Biloš 2

Scientific Reports volume 11 , Article number: 22516 ( 2021 ) Cite this article

21k Accesses

82 Citations

91 Altmetric

Metrics details

Cancer epidemiology
Climate change
Epidemiology
Lung cancer
Myocardial infarction
Risk factors

Air pollution is the fourth greatest overall risk factor for human health. Despite declining levels in Europe, air pollution still represents a major health and economic burden. We collected data from the Global Burden of Disease Study 2019 regarding overall, as well as ischemic heart disease (IHD), stroke, and tracheal, bronchus and lung cancer-specific disability adjusted life years (DALYs), years of life lost (YLL) and mortality attributable to air pollution for 43 European countries between 1990 and 2019. Concentrations of ambient particulate matter (aPM 2.5 ), ozone, and household air pollution from solid fuels were obtained from State of Global Air 2020. We analysed changes in air pollution parameters, as well as DALYs, YLL, and mortality related to air pollution, also taking into account gross national income (GNI) and socio-demographic index (SDI). Using a novel calculation, aPM 2.5 ratio (PMR) change and DALY rate ratio (DARR) change were used to assess each country’s ability to decrease its aPM 2.5 pollution and DALYs to at least the extent of the European median decrease within the analysed period. Finally, we created a multiple regression model for reliably predicting YLL using aPM 2.5 and household air pollution. The average annual population-weighted aPM 2.5 exposure in Europe in 1990 was 20.8 μg/m 3 (95% confidence interval (CI) 18.3–23.2), while in 2019 it was 33.7% lower at 13.8 μg/m 3 (95% CI 12.0–15.6). There were in total 368 006 estimated deaths in Europe in 2019 attributable to air pollution, a 42.4% decrease compared to 639 052 in 1990. The majority (90.4%) of all deaths were associated with aPM 2.5 . IHD was the primary cause of death making up 44.6% of all deaths attributable to air pollution. The age-standardised DALY rate and YLL rate attributable to air pollution were more than 60% lower in 2019 compared to 1990. There was a strong positive correlation (r = 0.911) between YLL rate and aPM 2.5 pollution in 2019 in Europe. Our multiple regression model predicts that for 10% increase in aPM 2.5 , YLL increases by 16.7%. Furthermore, 26 of 43 European countries had a positive DARR change. 31 of 43 European countries had a negative PMR change, thus not keeping up with the European median aPM 2.5 concentration decrease. When categorising countries by SDI and GNI, countries in the higher brackets had significantly lower aPM 2.5 concentration and DALY rate for IHD and stroke. Overall, air pollution levels, air pollution-related morbidity and mortality have decreased considerably in Europe in the last three decades. However, with the growing European population, air pollution remains an important public health and economic issue. Policies targeting air pollution reduction should continue to be strongly enforced to further reduce one of the greatest risk factors for human health.

Disparities in air pollution attributable mortality in the US population by race/ethnicity and sociodemographic factors

graphical representation of air pollution

Epidemiological trends and age-period-cohort effects on cardiovascular diseases burden attributable to ambient air pollution across BRICS

Global air pollution exposure and poverty

Introduction.

Clean air is considered one of the basic requirements of human health and well-being. However, more than 90% of the world population was exposed to air quality levels that exceeded the World Health Organization (WHO) Air Quality Guideline (AQG) limits in 2016 1 . Air pollution is the fourth greatest overall risk factor for human health globally, following high blood pressure, dietary risks, and smoking 2 . It has been associated with three of the leading causes of death in the world with significant shares of air pollution-related mortality: stroke (26%), ischemic heart disease (IHD) (20.2%), and primary cancer of the trachea, bronchus, and lung (TBL) (19%) 3 , 4 . Altogether, air pollution was linked to seven million deaths globally and in excess of 100 million disability-adjusted life years (DALYs) annually 5 , 6 . It also represents a major global annual economic impact of $5 trillion 7 .

Disparities in air quality have been observed based on a country's income. While air pollution in developed countries poses an important public health issue, it is even more pronounced in developing countries where fast-growing population along with widespread industrialization led to centers with poor air quality which became a serious threat to health 8 .

In Europe, emissions of air pollutants have been declining in the past decades 9 . Nevertheless, there were still more than 0.5 million deaths attributable to air pollution in 2013 while health-related external costs associated with air pollution reached close to €1 trillion annually in the European Union (EU) alone 10 , 11 .

In 2018, 73.6% of the EU urban population was exposed to excessive concentrations of particulate matter of diameter less than 2.5 microns (PM 2.5 ) which is considered the fifth leading mortality risk factor. The main contributors to the EU's PM 2.5 concentrations have been institutional, commercial, and household (55.5%), followed by road transport (10.7%) 12 . PM 2.5 pollution was associated with more than four million global deaths in 2016 with Europe counting for approximately 10% of that share 13 . Exposure to PM 2.5 pollution led to more than 1277 years of life lost (YLL) per 100,000 population in several European countries 14 . Long-term exposure to PM 2.5 pollution significantly increases both cardiopulmonary problems and lung cancer mortality, as well as risk for type two diabetes 15 , 16 . Conversely, one study showed that patients with lung cancer increased their life span by 0.35 years for every 10 μg/m 3 reduction of PM 2.5 concentration 17 . For short term exposure, every 10 μg/m 3 increase in PM 2.5 concentration was associated with 2.8% increase in PM-related mortality 18 .

Another major air pollutant is ozone, commonly found in urban areas which make up as much as 74.7% of the total EU population in 2019 19 . Long-term exposure to ozone has been linked to an increased risk of death from respiratory causes, as well as serious adverse pregnancy outcomes 20 , 21 .

The Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) is a multinational collaborative research study of disease burden that assesses mortality and disability from major diseases, injuries and risk factors, including air pollution 22 . The study is an ongoing effort and is designed to systematically incorporate information over time, and its latest iteration includes data from 1990 to 2019, by age and sex, and across more than 200 countries and territories. The study contains standard epidemiological measures such as incidence, prevalence, and death rates, as well as summary measures of health, such as DALYs.

Using the most recent 2019 GBD data, we did a comprehensive analysis of temporal and spatial trends for PM 2.5 and ozone concentrations in Europe from 1990 to 2019 with a focus on ambient PM 2.5 (aPM 2.5 ) concentrations due to its highest health impact. We also evaluated each country's ability to decrease its aPM 2.5 and DALY values to at least the extent of the European aPM 2.5 and DALY median reduction. Then, we analysed mortality, DALYs, and YLL attributable to air pollution for stroke, IHD, and TBL cancer, also taking into account socio-demographic index (SDI) and gross national income (GNI). This analysis could be used to raise awareness among policymakers to take action on this important public health issue which originates mainly from anthropogenic sources, and as such can be undone by precise and determined measures.

Sources of data

Data for 43 European countries regarding overall mortality, YLL, and DALYs attributable to air pollution by age, year, and sex were collected using the Global Health Data Exchange GBD Results Tool between 1990 and 2019. We also collected data for stroke, IHD and TBL cancer attributable to aPM 2.5 , household air pollution from solid fuels, and ozone. Detailed description of metrics, data sources, and statistics in GBD 2019 has been reported elsewhere 22 .

Estimates for exposure levels to aPM 2.5 and ozone were obtained from Health Effects Institute–State of Global Air 2020 where methods for assessing their exposure levels are also described 23 . In brief, exposure to aPM 2.5 was measured as the average annual PM 2.5 concentration in the air at a spatial resolution of a 0.1° × 0.1° grid cell, which reflects to 11 × 11 km at the equator. Ozone concentration is measured in parts per billion (ppb). Exposure to ozone was defined as the highest seasonal (six-month) average daily eight-hour maximum ozone concentration.

We used GBD 2019 estimations of deaths, YLLs, and DALYs attributable to air pollution. Relative risks were estimated based on meta-regression and systematic reviews done for GBD 2019. DALYs, YLLs, and attributable deaths were estimated by multiplying population attributable fractions (PAFs) by the relevant outcome quantity for each age, sex, location, and year. For continuous risk, PAFs are calculated using formula described in GBD 2019 study 22 . Inputs to estimation of PAFs for this study included continuous exposure distributions to air pollution, relative risk and the theoretical minimum risk exposure level (TMREL) for each group. TMREL was defined as the low point of the risk function and it represents the level of risk exposure that minimizes disease risk at the population level. Using PAF estimates, we calculated the number of deaths attributable to air pollution, DALYs and YLLs.

Data for SDI in 2019 for 43 European countries was obtained from the Global Burden of Disease Study 2019. For a robust analysis of disparities between groups in terms of aPM 2.5 pollution and disease burden, we divided these countries into 3 groups: high (> 0.850), medium (0.750–0.849), low (< 0.749) (Supplementary Table S4 ).

GNI per capita for 39 out of 43 European countries was obtained from the 2019 World Bank classification of world economies. Data for Andorra, Moldova, Monaco, and San Marino were not available 24 . The World Bank assigns the world’s economies into four income groups: low (< $1036), lower-middle ($1036–$4045), upper-middle ($4046–$12,535), and high-income (> $12,536) countries. Since 30 out of 39 analysed European countries are classified as high income, we divided those countries into an additional three groups: lower high income ($12,536–$36,766), moderate high income ($36,767–$60,997), very high income (> $60,997) for more robust analysis of disparities in aPM 2.5 pollution and disease burden between countries with different GNI (Supplementary Table S4 ). Also, since Ukraine is the only country in Europe classified as lower-middle income, we merged the lower-middle income and upper-middle income groups into one due to the same reasons.

Definitions

DALYs represent the overall number of years of potential life lost due to premature mortality and years of productive life lost due to disability and it is calculated as the sum of the aforementioned parameters. It summarizes the overall burden of disease and one DALY may be regarded as one year of healthy life lost 25 . DALYs can be expressed as the number of total DALYs or as DALY rate per 100,000 population. Additional methodologies for estimating DALYs have been described as part of GBD Study 2019 22 .

YLL is regarded as a summary measure of premature mortality. It estimates the years of potential life lost due to premature death, taking into account frequency of deaths and the age at which it occurs. YLLs were calculated by multiplying the estimated number of deaths by age with a standard life expectancy at that age. It can be expressed as a number of total YLLs or as YLL rate per 100,000 population. Additional methodologies for estimating YLLs are described in the GBD Study 2019 22 .

Death can be expressed as the rate per 100,000 population or as the total number of deaths. To calculate deaths attributable to air pollution, the total number of deaths is multiplied by the population attributable fraction (PAF), which may be interpreted as the proportion of deaths attributable to air pollution. Additional methodologies for estimating the number of deaths are described in the GBD Study 2019 22 .

SDI is a summary measure of socio‐demographic development status, strongly correlated with health outcomes. It is a geometric mean of the rankings of the lag-distributed income per capita, mean educational attainment for those age 15 or older, and fertility rate in those under 25 years old. It is expressed on a scale of 0 to 1, but for GBD 2019, values were multiplied by 100 for a scale of 0 to 100, where a location with an SDI of 0 has a theoretical minimum level of SDI, and a location with an SDI of 1 (prior to multiplying by 100 for reporting purposes) would have a theoretical maximum level of sociodemographic development relevant to health outcomes. Additional descriptions about SDI calculation can be found in the GBD Study 2019 22 .

GNI per capita represents the value produced by a country’s economy in a given year per one person, regardless of whether the value is produced domestically or abroad. Methodologies for calculation of GNI per capita in U.S. dollars are based on the Atlas method exchange rates described elsewhere 26 .

DALY rate ratio (DARR) represents the ratio between a country's DALY rate for a given year and a median DALY rate of all European countries for the same year. aPM 2.5 ratio (PMR) represents the ratio between a country's aPM 2.5 concentration for a given year and a median aPM 2.5 level of all European countries for the same year.

YLL rate ratio (YRR) represents the ratio between a country's YLL rate for a given year and the median YLL rate of all European countries for the same year.

Death rate ratio (DRR) represents the ratio between a country's death rate for a given year and the median death rate of all European countries for the same year.

Statistical analyses

Burden of disease (e.g. attributable DALYs or mortality) calculation requires a few factors to be taken into account: spatial and temporal estimates of population-weighted exposure, TMREL, estimation of relative risk from exposure, as well as estimates of DALYs and deaths for diseases linked causally to air pollution. First, the data for relative risk and estimates of exposure of the population are combined which allows for the calculation of PAF, a proportion of DALYs and deaths in a population that can be attributed to exposure (e.g. to air pollution) above TMREL. Finally, the number of DALYs and deaths for certain diseases are multiplied by PAF and the end value gives an estimation of the burden attributable to the exposure. Specifically, we used DALYs, mortality and YLL attributable to air pollution overall, as well as to aPM 2.5 , household air pollution and ozone. A more detailed description of these methods can be found in GBD Study 2015 and 2019 6 , 22 .

Since the primary and final time points of our study were 1990 and 2019, we determined DARR and PMR for each European country for both years. We then defined change in DARR and PMR between those two time points as a new variable (DARR change and PMR change) and used it to quantify each country's ability to decrease its aPM 2.5 and DALY values to at least the extent of the European aPM 2.5 concentration and DALY median decrease between those two temporal points. Furthermore, 1990 and 2019 also respectively represent the largest and smallest value of European median value for aPM 2.5 rate and DALY rate with other values following a linear decrease during that 29-year period, starting from 1990 (Supplementary Figure S1 ).

If DARR change or PMR change are positive, a country shows a reduction in its aPM 2.5 or DALY values minimally to the extent of the European aPM 2.5 concentration or DALY median decrease, but if DARR change or PMR change are negative, a country cannot follow the European median reduction. However, countries which improved their own DALY rate and aPM 2.5 concentration may still be represented with a negative DARR change or PMR change if that improvement is lower than the extent of the European DALY or aPM 2.5 concentration median decrease. Using this formula, we also calculated YRR and DRR, as well as their change between 1990 and 2019. Other variables in different time points can also be analysed in a similar manner. A visual representation of the calculation can be found in Supplementary Figure S6 .

For each country, all variables are represented as a numerical value along with a 95% uncertainty interval (95% UI). For descriptive analysis of subgroups and all countries together, median and interquartile range (IQR) are predominantly used due to significant dataset variability and deviation from Gaussian distribution, unless stated otherwise. On the other hand, the death number is the only variable presented as a cumulative death number of all countries within a certain subgroup.

Due to significant dataset variability and deviation from Gaussian distribution, dependence between variables (YLL and aPM 2.5 concentration, PMR and DARR) was established using Spearman correlation test. Furthermore, Kruskal–Wallis test by ranks with Dunn’s multiple comparisons test was used to determine statistical significance between subgroups of countries classified by their GNI and SDI.

Multiple linear regression was used to predict the outcome of YLL rate attributable to air pollution for 2019 using air pollution-related explanatory variables and to further establish relationship between air pollution parameters and health outcomes. Explanatory variables used in the model were aPM 2.5 and household air pollution (HAP) for 2019. Both explanatory and response variables were log-transformed (ln-transformed). Ozone was excluded from the model because it failed to meet the assumption of linear relationship with the response variable. Even after multiple data transformations, it seemed that ozone did not have any significant relationship with the response variable. Furthermore, forcing the ozone into the model did not produce any significant improvement in proportion of explained variance. Alpha value for all statistical tests was set at 0.05. Data was analysed using GraphPad Prism 9 and Statistica 13.5.

Overall change of air quality parameters and morbidity and mortality estimates attributable to air pollution in Europe from 1990 to 2019

The average aPM 2.5 exposure of European countries was 20.8 μg/m 3 (95% CI 18.3–23.2) in 1990, while in 2019 it was 33.7% lower at 13.8 μg/m 3 (95% CI 12.0–15.6). All European countries reduced their annual population-weighted aPM 2.5 concentration, except Monaco (Fig. 1 , Supplementary Table S1 ). Western, Nordic and Baltic countries showed the biggest improvement in general, whereas progress was smaller for Eastern and Southeastern countries. On the other hand, average seasonal population-weighted ozone concentrations did not reduce as much as aPM 2.5 with 44.7 ppb, a 6.5% decrease compared to 41.8 ppb in 2019 (Supplementary Figure S2 ).

The average annual population-weighted aPM 2.5 concentration in European countries for 1990 ( a ) and 2019 ( b ). Countries are categorised by parameters of the European Environment Agency for 2019 into groups based on aPM 2.5 concentrations [μg/m 3 ] in 2019: good (0–10.0), fair (10.0–19.9), moderate (20.0–24.9) and poor (≥ 25.0). Percent reduction in annual population-weighted aPM 2.5 concentration in 2019 compared to 1990 ( c ). The figure was made in Adobe Illustrator (version 24.1., URL: https://www.adobe.com/products/illustrator.html ).

We then analysed death number, death rates and DALY rates attributable to air pollution for 2019 in all 43 countries (Table 1 ). There were in total 368,006 deaths in Europe attributable to air pollution, a reduction of 271 046 (42.4%) compared to 1990 with 37 countries lowering their number of deaths (Fig. 2 ). In the same time period, an overall 43.9% decrease in total number of deaths attributable to air pollution was observed in the subset of EU countries. Estonia had the most significant decrease in mortality attributable to air pollution with a 82.3% reduction in 2019 when compared to 1990. It is followed by Norway and Sweden with 73.5% and 72.8% fewer deaths, respectively. During the 29-year time period, we also observed a decreasing trend of all-cause median death rate of all European countries, with a total reduction of 40.6% (Fig. 3 ). EU countries had a 1.3% lower death rate and 5.6% lower DALY rate attributable to air pollution (Supplementary Table S2 ) compared to all European countries. Furthermore, reduction of death rate attributable to air pollution was more pronounced when compared to overall death rate. It was 50.0 in 1990, whereas in 2019 it was 16.7, a 66.6% decrease.

Percent change in number of deaths attributable to air pollution in 2019 compared to 1990.

Median European death rate from 1990 to 2019.

IHD was the primary cause of death in Europe making up 44.6% of all deaths attributable to air pollution. Stroke and TBL cancer had a smaller contribution to the total death number with 25.2% and 10.7%, respectively. When analysing the air pollution parameters, the majority (90.4%) of all deaths were associated with aPM 2.5 pollution. Therefore, due to the high share in total number of deaths among all air pollutants, we primarily focused on analysing effects of aPM 2.5 on health.

Since we showed that aPM 2.5 concentration in Europe decreased, we aimed to determine if the contribution of death rate attributable to air pollution in all-cause death rate also decreased. In 1990, it was 6.3% while in 2019 it was 3.5%, a reduction of 44.4%. We then aimed to explore how air pollution impacts premature mortality. Our analysis showed that a total of 24,917 years of life were lost per 100,000 population in Europe in 2019 due to health conditions associated with air pollution exposure (Fig. 4 ). This is a 63% decrease compared to 1990 (YLL = 67 258). Among all European countries in 2019, Finland had both the lowest YLL rate (60.9) and aPM 2.5 concentration (5.6 μg/m 3 ), while North Macedonia had the highest YLL rate (2214.9) and aPM 2.5 concentration (30.3 μg/m 3 ). Furthermore, a strong positive correlation was observed between aPM 2.5 concentration and YLL rate (r = 0.911, p < 0.0001) (Supplementary Figure S3 ).

Age-standardised YLL rate per 100,000 and average annual population-weighted aPM 2.5 in 2019.

With IHD being the primary cause of death attributable to air pollution in Europe, we analysed YLL rate among European countries and observed that IHD also contributed the most with 41.2% of total YLL rate attributable to air pollution in 2019. IHD contributed the most to the YLL rate in Belarus (63.6%), while in Denmark it had the lowest contribution (22.9%).

Progress of each country compared to Europe overall in terms of DALYs and aPM 2.5

To evaluate progress and the dynamics of reduction of air pollution parameters, we aimed to determine how efficient each European country was in reducing its DALY and aPM 2.5 values to at least the extent of the European median decrease from 1990 to 2019. Therefore, European countries were compared using the DARR change attributable to air pollution in the first and final year of the analysed period. This value represents a difference between each country's DALY rate attributable to air pollution for the two given years (1990 and 2019) divided by the median European DALY for 1990. Figure 5 shows that 26 of 43 European countries improved their DARR, i.e. had a positive DARR change. These countries decreased their DALY rate attributable to air pollution minimally to the extent of the European median reduction. Estonia and Finland had the greatest improvement by decreasing DARR from 1990 to 2019 by 64.5% and 51.2%, respectively. On the other hand, 17 of 43 countries increased their DARR and are represented with a negative DARR change. Therefore, they have not been able to reduce their DALY values to at least the extent of the European median decrease. Monaco and Montenegro had the most negative values with an increase of 146.0% and 79.5%, respectively.

DARR change (%) and PMR change (%) for European countries.

PMR was also determined for each European country. Similarly to DARR change, PMR change was used as a measure of each country's ability to decrease its aPM 2.5 value minimally to the extent of the European median reduction. Although all countries (except Monaco) decreased their aPM 2.5 level in 2019 compared to 1990, 31 of 43 European countries had a negative PMR change and an unfavorable trend of increasing PMR in 2019 compared to 1990, thus not reducing its aPM 2.5 concentration to at least the extent of decrease in the European median. Monaco had the most prominent PMR change with an increase in PMR of 92% while other countries showed an increase of less than 50%. Although Finland increased its PMR by 7%, it still remained the country with the lowest aPM 2.5 level among European countries in 2019 and second lowest in 1990. Italy and Netherlands had a neutral PMR change (0%) because they had equal PMRs both in 1990 and 2019. 10 of 43 countries had a positive PMR change and favorable trend of decreasing aPM 2.5 values to at least the extent of the European median decrease with Switzerland having the largest reduction in PMR of 8%, followed by Norway and Denmark with a decrease of 6.9% and 5.9%, respectively. Furthermore, using the Spearman test, positive correlation between DARR and PMR was established both for 1990 (r = 0.854) and 2019 (r = 0.921). Using this calculation, YRR and DRR were also analysed (Supplementary Table S3 ).

In order to explore whether aPM 2.5 and HAP can significantly predict YLL attributable to air pollution for 2019, a multiple regression was performed. Our model statistically significantly predicted YLL values and also explained a significant proportion of variance in YLL (R 2 = 0.885, F(2,40) = 154.116, p < 0.0001). Both aPM 2.5 (b = 1.623, p < 0.0001) and HAP (b = 0.150, p < 0.0001) were statistically significant predictors. The final model was:

As a practical example of this model, it predicts that for a 10% increase in aPM 2.5 , YLL increases by 16.7%. Furthermore, for a 10% increase in household air pollution, YLL increases by 1.4%.

Impact of economic and social factors on air pollution-related disease burden

Then, we aimed to determine if there was a significant difference in aPM 2.5 concentration and DALY rate attributable to air pollution based on a country's SDI. Thus, countries were grouped by their SDI in three categories (Fig. 6 ). Kruskal–Wallis test showed significant differences between groups for all three diseases: IHD (H(2) = 22.344, p < 0.0001), stroke (H(2) = 21.847, p < 0.0001) and TBL cancer (H(2) = 8.258, p = 0.016). When analysing differences between groups for IHD and stroke attributable to air pollution in 2019, low SDI countries showed more than a 11-fold higher DALY rate both for IHD and stroke when compared to high SDI countries. Furthermore, more than a fivefold higher DALY rate for IHD and nearly fourfold higher DALY rate for stroke was observed in medium SDI countries compared to high SDI countries. When looking at TBL cancer, DALY rate showed lower variability among groups, but a statistically significant difference ( p = 0.028) in DALY rate was observed when comparing high SDI to medium SDI countries. Significant differences (H(2) = 16.844, p = 0.0002) were also observed between groups of countries when comparing their aPM 2.5 concentrations (Supplementary Figure S4B ).

Comparison of countries by SDI and DALY rate attributable to air pollution for IHD, stroke and TBL cancer in 2019. Countries are categorised by SDI into three categories: high SDI, medium SDI and low SDI. Groups are represented as median with interquartile range. Statistical analysis by Kruskal–Wallis test by ranks with Dunn’s multiple comparisons test. Bonferroni correction for multiple tests was used to adjust significance values.

In addition, we also compared groups of countries based on their GNI to assess if a country’s income might be a differentiating factor in terms of DALY rate attributable to air pollution. There were significant differences between groups for all three diseases: IHD (H(3) = 28.038, p < 0.0001), stroke (H(3) = 28.963, p < 0.0001) and TBL cancer (H(3) = 15.550, p = 0.001). We showed that groups of countries with lower GNI had significantly higher DALY compared to groups of countries with higher GNI (Supplementary Figure S4A ). This effect is most prominent when looking at countries in the lowest economic bracket which have more than a 11-fold higher DALY rate for IHD and nearly 25-fold higher DALY rate for stroke when compared to very high income (VHI) countries. Furthermore, significant differences (H(3) = 19.918, p = 0.0002) were also observed between GNI groups when comparing their aPM 2.5 concentrations (Supplementary Figure S4B ). These findings align well with SDI comparisons, aided by a strong correlation between GNI and SDI (Spearman r = 0.893, p < 0.0001). Taken together, these comparisons show that countries with higher GNI and SDI have lower DALY rates attributable to air pollution compared to countries in lower GNI and SDI brackets.

Air pollution is the most important environmental risk to human health 27 . It is also perceived among Europeans as the second biggest environmental concern 28 . A growing interest in the topic of air pollution has led to public and political actions ultimately successfully reducing air pollution levels.

We showed that all European countries excluding Monaco decreased the concentration of the leading air pollutant aPM 2.5 by up to 44% in 2019 compared to 1990. One of the reasons that Monaco is an exception to this positive trend might be the very high population density (more than 170-fold more than the average population density of the European Union) which influences the aPM 2.5 value since the aPM 2.5 exposure calculation accounts for population density. The overall decrease in Europe may partly be due to strong and proactive political legislations which have proved effective in curbing air pollution in many studies 29 , 30 , 31 . The Convention on Long-range Transboundary Air Pollution has successfully reduced air pollution in Europe, whereas the European Green Deal focuses on making Europe climate neutral by 2050 32 . The act empowers to prioritize a sustainable industry, energy efficiency by using clean energy, the importance of recycling, optimizing agriculture, and sustainable mobility 33 . Reductions in air pollution were closely followed by reduced DALYs and mortality attributable to air pollution improving the overall population health (Supplementary Figure S5 ). However, despite the progress made, air pollution remains an important pan-European public health issue with nearly ¾ of European countries still exceeding the annual WHO AQG for aPM 2.5 pollution of 10 μg/m 3 in 2019. This is especially important in the context of the ever-growing European population where urbanization is expected to increase from 74.7% to approximately 83.7% in 2050 19 , 34 .

Study from Boldo et al. with 23 European cities showed that life expectancy at age 30 would increase by a range between one month and more than two years if long-term exposure to aPM 2.5 level was reduced to 15 μg/m 3 35 . Also, a recent study from almost 1000 European cities calculated that 51,213 deaths per year could be avoided if PM 2.5 exposure was compliant with WHO air pollution guidelines 36 .

As a result of improved air quality, more than 85% of countries in Europe had a lower number of deaths attributable to air pollution in 2019 compared to 1990. Also, the overall number of deaths attributable to air pollution was lowered by more than 270,000 (42.4%). Interestingly, the share of death rate attributable to air pollution in the overall all-cause death rate in Europe decreased more than the overall death rate, thus outpacing it from an approximate 6% share in 1990 to nearly 3.5% in 2019. On the other hand, despite reducing air pollution, Southeast European countries Bosnia and Herzegovina and North Macedonia still had the highest aPM 2.5 levels in Europe in 2019, three times higher than the WHO AQG. Our findings align with Lelieveld et al. who also found that air pollution-related mortality per capita was high in Eastern European countries, especially concerning cardiovascular mortality 37 . Death rates in Bosnia and Herzegovina and North Macedonia were five-fold and seven-fold higher than the European median, respectively. Their DALY rates were also the highest in Europe, multiple times higher than both the EU and European median and up to 32 times higher than Iceland, which had the lowest DALY rate, thus indicating room for improvement. In another study, Lehtomäki et al. showed that Iceland had the lowest death rate among five Nordic countries, all of which have relatively low levels of air pollution, and generally meet the EU guideline values 38 . Furthermore, for a deeper understanding of time trends in burden estimates, a decomposition of total changes in DALY rates or mortality over time for each European country could be performed, taking into account the contribution of population size change, age, cause specific-mortality rates (excluding the effect of air pollution), and air pollution exposure. These analyses could provide a useful overview of how each factor contributes to changes in DALY rates or mortality, and thus inform policy makers, as well as health officials, about potentially implementing specific measures to address the factors contributing most to disease burden. Decomposition methods have been previously reported in GBD studies, as well as papers that focused primarily on this method, which due to its significant comprehensiveness, was not within the scope of our research 6 , 39 , 40 .

As a summary measure of premature mortality, YLL has been highly associated with air pollution and each year more than 200 million years of life are lost due to air pollution globally 41 . Our results align closely since we showed a strong positive association between aPM 2.5 concentration and YLL. One study using data from 72 Chinese cities estimated that for every 10 μg/m 3 increase in PM 2.5 an additional 0.43 years of life are lost, whereas PM 2.5 levels in accordance with WHO AQG would result in 0.14 years of gain in life expectancy 42 . To get a better and more comprehensive understanding of the relationship between disease burden metrics such as YLL and aPM 2.5 and household air pollution, the model we created came very useful. Due to its significant prediction of YLL attributable to air pollution using aPM 2.5 and household air pollution, we think this might be of good use to policymakers and researchers in the field to reliably predict YLL based on trends of air pollution components, especially during a period of years. Furthermore, since the model predicts that a 10% increase in aPM 2.5 would result in YLL increase of 16.7%, whereas a 10% increase in household air pollution would increase YLL by only 1.4%, it strongly supports that aPM 2.5 has the predominant impact on disease burden, thus pointing to the importance of curbing this air pollution parameter.

Comparing the results to the overall trend in the region is also a significant factor regarding not only a country’s environmental consciousness, but also the ability to leverage resources to combat air pollution. Both PMR change and DARR change are important to see if the country is reducing its aPM 2.5 concentration and DALY values at least at the pace of the European region between 1990 and 2019. Despite almost all countries decreasing their aPM 2.5 concentrations in 2019 , only 10 countries actually did that to at least the extent of the European median decrease. Sweden and Norway were among the countries with the lowest aPM 2.5 concentrations in 1990, yet still managed to decrease their aPM 2.5 values in 2019 more than the European median, thus having a positive PMR change. On the other hand, Bosnia and Herzegovina and Albania had among the most negative PMR changes in spite of reducing their aPM 2.5 concentrations. Belis et al. identified energy production in inefficient coal-fueled power plants as one of the main sources of PM 2.5 in the Western Balkans. Also, agriculture and residential combustion significantly affected PM 2.5 levels 43 . Since some Western Balkan countries had among the highest aPM 2.5 levels both in 1990 and 2019, more powerful ways of curbing air pollution are needed.

Similarly, we used DARR change to evaluate each country's progress in terms of reduction in DALY rates attributable to air pollution compared to the European median change. All 43 countries reduced their own DALY rate in 2019 compared to 1990, but less than 2/3 had a decrease to at least the extent of the European median reduction, thus having a positive DARR change. EU countries Finland, Sweden, and Estonia all had DARR change above 40%, while Monaco, Ukraine and Montenegro had the most negative DARR change, up to -146%. The wide disparity in terms of DARR change may be due to more strict environmental policies in EU countries, ambitious targets for emission reductions, and economic power and development 44 . Therefore, both aPM 2.5 concentration and DALY rate, as well as DARR change and PMR change are complementary methods which should be used when evaluating changes in aPM 2.5 concentration and DALY rate between countries. It is important to note that multiple diverse variables and geographical regions could used in this calculation, which gives it a breath of flexibility and applicability in various scenarios.

Taking into account social and economic factors using SDI and GNI in regard to air pollution-related health burden is an important metric. We showed that countries in the lowest SDI category in Europe had a higher aPM 2.5 concentration compared to those in the highest. This aligns well with another study showing that the country's lower development status might be associated with overall poorer air quality 8 . Developing countries undergoing the process of intense urbanization and industrialization became the countries with the largest air pollution-related burdens in recent years. Furthermore, indoor air pollution originating from coal and biomass in the form of wood, dung and crop residues for domestic energy represents a major environmental and public health challenge in developing economies, especially in rural areas 45 . Even though we expected more developed countries to have lower air pollution parameters, we were surprised by how large some of the differences were when comparing countries. Low SDI countries had more than a 11-fold higher DALY rate both for IHD and stroke when compared to high SDI countries. This difference has even greater meaning when taking into account that stroke and IHD are two of the leading causes of death in the world. Similar striking differences were found when comparing countries by GNI, thus excluding educational attainment and total fertility rate which are part of SDI. Upper middle income (UMI) countries had more than a 11-fold higher DALY rate for IHD and nearly 25-fold higher DALY rate for stroke when compared to VHI countries. This big difference emphasizes a larger than expected gap between more and less developed countries in Europe and shows how disparities in controlling air pollution can negatively affect population health. This is a strong call to action, especially in lower developed countries, to double down on curbing air pollution considering that stroke, IHD and TBL cancer made up more than 80% of all European deaths attributable to air pollution in 2019 alone. The difference in economic power and DALYs might also be due to other factors such as a more comprehensive and stronger healthcare system in wealthier countries. A study showed that better and more extensive quality of healthcare is needed to improve patient outcomes, especially since 63.8% of deaths in Eastern Europe occurred due to use of poor-quality services 46 . The somewhat smaller statistical significance for TBL cancer might be due to cancer being a complex heterogeneous disease which can be attributable to both genetics and lifestyle and thus various factors might be triggering its genesis in different and unequal manners 47 .

With cooperation at intra- and inter-national levels, strong policies could be implemented at curbing both air pollution and premature mortality and morbidity, while serving as a catalyst for economic development and promotion of healthy lifestyle 48 . Emissions from vehicles could be reduced by prioritizing green and sustainable forms of transport such as rapid and optimized urban and international transport, cycling, as well as implementing stricter vehicle emissions standards and working on more efficient engine technologies 48 , 49 . Making cities more compact and with energy efficient homes, optimizing urban transport and waste management will be of utmost importance to mitigate air pollution increase. Improving the management of agricultural waste and livestock manure, while reducing agriculture field burning and promoting healthy diets low in processed meat and rich in plant-based food will keep the food production environmentally sustainable 50 . Despite improvements in wastewater management and pollution abatement technology, industry is still a significant source of pollutant releases in Europe 51 . Further implementation of clean technologies, filters, and recovery of gas released during fossil fuel production are recommended to optimize soil, water and air quality management 52 . Public actions and national robust policies could have a long-lasting impact on bringing down air pollution, as well as growing human health and welfare.

Our study has several limitations. First, important causes of death like chronic obstructive pulmonary disease, dementias, diabetes and kidney disease were not individually addressed in the context of air pollution-related morbidity and mortality. Second, DARR change and PMR change might not give a clear indication of the country's progress in certain conditions. A negative DARR change and/or PMR change might underestimate the country's progress since it could already have low DALY rates or aPM 2.5 values. Third, estimates on levels of aPM 2.5 and ozone might be skewed given the smaller numbers and less well diversified locations of air quality monitoring stations, as well as the spread of air pollution from other countries through changes in wind pattern with temperature 53 , 54 . Also, GBD study estimates of household air pollution include solid fuels used for cooking, but not for heating. Fourth, availability of primary data is a major limitation of the GBD study and as such applies here for mortality and morbidity estimates, along with other general limitations described in the GBD Study 2019 22 . Fifth, decomposing the total changes in mortality or DALY rates over time for each European country, taking into account the contribution of population size change, age, cause specific-mortality rates (excluding the effect of air pollution), and air pollution exposure was not performed which could have given the paper a deeper understanding of the factors influencing them the most. Sixth, recommendations on how to reduce air pollution might not be feasible for every country in the same way due to different dominant industries and economic power. Finally, air pollution, as well as mortality and morbidity estimates for each country might not be representative for all the country's regions. Also, specific air pollution-related medical conditions might not be represented in an equal manner in the whole population and certain subpopulations might be more or less affected. Even with these limitations, our study provides a useful overview and analysis regarding the health effects of air pollution in Europe using the most recent data available.

In conclusion, Europe made significant progress in decreasing aPM 2.5 concentration, mortality and disease burden attributable to air pollution in the last three decades. However, nearly 75% of Europeans still live in areas where aPM 2.5 concentration do not meet WHO AQG. Even though implementing air pollution reduction measures may be a significant challenge for some countries, with population growth and increased urbanization in Europe, air quality should be prioritized for long term economic growth and improved overall population health.

Data availability

The dataset regarding overall, as well as ischemic heart disease (IHD), stroke, and tracheal, bronchus and lung cancer-specific disability adjusted life years (DALYs), years of life lost (YLL) and mortality attributable to air pollution for 43 European countries between 1990 and 2019 are available in the Global Burden of Disease 2019 repository which can be found on this web page: http://www.healthdata.org/gbd/2019 . Socio-demographic index for each country can be found in the same database. The dataset regarding concentrations of ambient particulate matter less than 2.5 microns in size, ozone, and household air pollution from solid fuels were obtained from State of Global Air 2020 which can be found on this web page: https://www.stateofglobalair.org/ . The dataset regarding the gross national income of each European country was obtained from the World Bank Classification and can be found on this web page: https://data.worldbank.org/indicator/NY.GNP.PCAP.CD . Other data used in this manuscript can be found within the reference section.

Ambient (outdoor) air pollution. World Health Organization. https://www.who.int/news-room/fact-sheets/detail/ambient-(outdoor)-air-quality-and-health (2018).

Rafaj, P. et al. Outlook for clean air in the context of sustainable development goals. Glob. Environ. Change 53 , 1–11 (2018).

Article Google Scholar

The top 10 causes of death. World Health Organization. https://www.who.int/news-room/fact-sheets/detail/the-top-10-causes-of-death (2020).

Global Burden of Disease Collaborative Network. Global Burden of Disease Study 2019 (GBD 2019) Results. Institute for Health Metrics and Evaluation. http://ghdx.healthdata.org/gbd-results-tool (2020).

Air Pollution. World Health Organization. https://www.who.int/health-topics/air-pollution#tab=tab_1 (2020).

Cohen, A. J. et al. Estimates and 25-year trends of the global burden of disease attributable to ambient air pollution: an analysis of data from the Global Burden of Diseases Study 2015. Lancet 389 , 10082. https://doi.org/10.1016/S0140-6736(17)30505-6 (2017).

World Bank; Institute for Health Metrics and Evaluation. The cost of air pollution: Strengthening the economic case for action. World Bank. https://openknowledge.worldbank.org/handle/10986/25013 (2016).

Mannucci, P. M. & Franchini, M. Health effects of ambient air pollution in developing countries. Int. J. Environ. Res. Public Health 14 , 1048. https://doi.org/10.3390/ijerph14091048 (2017).

Article CAS PubMed Central Google Scholar

Air pollution. European Environment Agency. https://www.eea.europa.eu/themes/air/intro (2020)

European Environment Agency. Air quality in Europe: 2016 Report. European Environment Agency. https://www.eea.europa.eu/publications/air-quality-in-europe-2016 (2016).

Commission Staff Working Document — Impact Assessment. European Commission. http://ec.europa.eu/environment/archives/air/pdf/Impact_assessment_en.pdf (2013).

Air Quality in Europe: 2018 Report. European Environment Agency. https://www.eea.europa.eu/publications/air-quality-in-europe-2018 (2018).

Anenberg, S. C. et al. Particulate matter-attributable mortality and relationships with carbon dioxide in 250 urban areas worldwide. Sci. Rep. 9 , 11552. https://doi.org/10.1038/s41598-019-48057-9 (2019).

Article ADS CAS PubMed PubMed Central Google Scholar

The European environment — state and outlook 2020: Knowledge for transition to a sustainable Europe. European Environment Agency. https://www.eea.europa.eu/publications/soer-2020 (2019).

Xu, X. et al. Long-term exposure to ambient fine particulate pollution induces insulin resistance and mitochondrial alteration in adipose tissue. Toxicol. Sci. 124 , 99–98 (2011).

Xing, Y. F., Xu, Y. H., Shi, M. H. & Lian, Y. X. The impact of PM2.5 on the human respiratory system. J. Thorac. Dis. 8 , E69–E74 (2016).

Google Scholar

Correia, A. W. et al. Effect of air pollution control on life expectancy in the United States: An analysis of 545 U.S. counties for the period from 2000 to 2007. Epidemiology 24 , 23–31 (2013).

Kloog, I., Ridgway, B., Koutrakis, P., Coull, B. A. & Schwartz, J. D. Long- and short-term exposure to PM2.5 and mortality: Using novel exposure models. Epidemiology 24 , 555–561 (2013).

Urban population (% of total population) — European Union. World Bank. https://data.worldbank.org/indicator/SP.URB.TOTL.IN.ZS?locations=EU (2018).

Jerrett, M. et al. Long-term ozone exposure and mortality. N. Engl. J. Med. 360 , 1085–1095 (2009).

Article CAS Google Scholar

Bekkar, B., Pacheco, S., Basu, R. & DeNicola, N. Association of air pollution and heat exposure with preterm birth, low birth weight, and stillbirth in the US: A systematic review. JAMA Netw. Open. 3 , e208243. https://doi.org/10.1001/jamanetworkopen.2020.8243 (2020).

Article PubMed PubMed Central Google Scholar

GBD 2019 Diseases and Injuries Collaborators. Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: A systematic analysis for the Global Burden of Disease Study 2019. Lancet . 396, 1204–1222 (2020).

Health Effects Institute. State of Global Air 2020. Health Effects Institute. www.stateofglobalair.org (2020).

World Development Indicators (Series: GNI per capita, Atlas method (current US$). The World Bank. http://data.worldbank.org/indicator (2020).

Global Burden of Disease Cancer Collaboration. Global, regional, and national cancer incidence, mortality, years of life lost, years lived with disability, and disability-adjusted life-years for 32 cancer groups, 1990 to 2015: A systematic analysis for the global burden of disease study. JAMA Oncol. 3 , 524–548 (2017).

The World Bank Atlas method — detailed methodology. World Bank. https://datahelpdesk.worldbank.org/knowledgebase/articles/378832-what-is-the-world-bank-atlas-method (2020).

Chau, T. T. & Wang, K. Y. An association between air pollution and daily most frequently visits of eighteen outpatient diseases in an industrial city. Sci. Rep. 10 , 2321. https://doi.org/10.1038/s41598-020-58721-0 (2020).

Special Eurobarometer 468: Attitudes of European citizens towards the environment. EU Open Data Portal . https://data.europa.eu/euodp/en/data/dataset/S2156_88_1_468_ENG (2021).

Turnock, S. T. et al. The impact of European legislative and technology measures to reduce air pollutants on air quality, human health and climate. Environ. Res. Lett. 11 , 024010. https://doi.org/10.1088/1748-9326/11/2/024010 (2016).

Article ADS CAS Google Scholar

Guerreiro, C. B. B., Foltescu, V. & de Leeuw, F. Air quality status and trends in Europe. Atmos. Environ. 98 , 376–384 (2014).

Kuklinska, K., Wolska, L. & Namiesnik, J. Air quality policy in the U.S. and the EU—A review. Atmos. Pollut. Res. 6 , 129–137 (2015).

Byrne, A. The 1979 convention on long-range transboundary air pollution: Assessing its effectiveness as a multilateral environmental regime after 35 years. Trans. Environ. Law 4 , 37–67 (2015).

Communication from the Commission to the European Parliament, the European Council, The Council, the European Economic and Social Committee and the Committee of the Regions: The European Green Deal. European Commission. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=COM%3A2019%3A640%3AFIN (2019).

Developments and Forecasts on Continuing Urbanisation. European Commission. https://knowledge4policy.ec.europa.eu/foresight/topic/continuing-urbanisation/developments-and-forecasts-on-continuing-urbanisation_en (2021).

Boldo, E. et al. Apheis: Health impact assessment of long-term exposure to PM 2.5 in 23 European cities. Eur. J. Epidemiol. 21 , 449–458 (2006).

Khomenko, S. et al. Premature mortality due to air pollution in European cities: A health impact assessment. Lancet Planet. Health 5 , e121–e134 (2021).

Lelieveld, J., Klingmüller, K. & Pozzer, A. Cardiovascular disease burden from ambient air pollution in Europe reassessed using novel hazard ratio functions. Eur. Heart J. 40 , 1590–1596 (2019).

Lehtomäki, H., Geels, C. & Brandt, J. Deaths attributable to air pollution in nordic countries: Disparities in the estimates. Atmosphere 11 , 467 (2020).

Article ADS Google Scholar

Cheng, X. Population ageing and mortality during 1990–2017: A global decomposition analysis. PLoS Med. 17 , 6. https://doi.org/10.1371/journal.pmed.1003138 (2020).

Ščasný, M., Ang, B. W. & Rečkaa, L. Decomposition analysis of air pollutants during the transition and post-transition periods in the Czech Republic. Renew. Sustain. Energy Rev. 145 , 111137. https://doi.org/10.1016/j.rser.2021.111137 (2021).

Lelieveld, J. et al. Loss of life expectancy from air pollution compared to other risk factors: A worldwide perspective. Cardiovasc. Res. 116 , 1910–1917 (2020).

Qi, J. et al. Potential gains in life expectancy by attaining daily ambient fine particulate matter pollution standards in mainland China: A modeling study based on nationwide data. PLoS Med. 17 , e1003027. https://doi.org/10.1371/journal.pmed.1003027 (2020).

Belis, C. A. et al. Urban pollution in the Danube and Western Balkans regions: The impact of major PM2.5 sources. Environ. Int. 133 , 105158. https://doi.org/10.1016/j.envint.2019.105158 (2019).

Article CAS PubMed PubMed Central Google Scholar

Koolen, C. D. & Rothenberg, G. Air pollution in Europe. Chemsuschem 12 , 164–172 (2019).

Bruce, N., Perez-Padilla, R. & Albalak, R. Indoor air pollution in developing countries: A major environmental and public health challenge. World Health Organ. 78 , 1078–1092 (2000).

CAS Google Scholar

Kruk, M. E. et al. Mortality due to low-quality health systems in the universal health coverage era: A systematic analysis of amenable deaths in 137 countries. Lancet 392 , 2146–2147 (2018).

Chen, Z., Fillmore, C. M., Hammerman, P. S., Kim, C. F. & Wong, K.-K. Non-small-cell lung cancers: A heterogeneous set of diseases. Nat. Rev. Cancer 14 , 535–546 (2014).

Scovronick, N. Reducing Global Health Risks Through Mitigation of Short-Lived Climate Pollutants . (World Health Organization, 2015).

Overview of Air Pollution from Transportation. United States Environmental Protection Agency. https://www.epa.gov/transportation-air-pollution-and-climate-change/learn-about-air-pollution-transportation (2021).

Interventions: Agriculture. World Health Organization. https://www.who.int/airpollution/ambient/interventions/agriculture/en/ (2021).

Air Quality in Europe: 2020 Report. European Environment Agency. https://www.eea.europa.eu//publications/air-quality-in-europe-2020-report (2020).

Industrial pollution in Europe. European Environment Agency . https://www.eea.europa.eu/data-and-maps/indicators/industrial-pollution-in-europe-3/assessment (2021).

De Sario, M., Katsouyanni, K. & Michelozzi, P. Climate change, extreme weather events, air pollution and respiratory health in Europe. Eur. Respir. J. 42 , 826–843 (2013).

Kracht, O. et al. Spatial representativeness of air quality monitoring sites — Outcomes of the FAIRMODE/AQUILA intercomparison exercise. European Commission - Joint Research Centre. https://publications.jrc.ec.europa.eu/repository/handle/JRC108791 (2017).

Download references

Acknowledgements

V.B. made in Adobe Illustrator (version 24.1., URL: https://www.adobe.com/products/illustrator.html ) the maps used in Figure 1 , Supplementary Figure S1 , Supplementary Figure S2 and Supplementary Figure S6 .

No funding was provided for this research.

Author information

Authors and affiliations.

Department of Neurobiology, Harvard Medical School, 220 Longwood Avenue, Boston, MA, 02115, USA

Alen Juginović

School of Medicine, University of Split, Šoltanska 2, 21 000, Split, Croatia

Miro Vuković, Ivan Aranza & Valentina Biloš

You can also search for this author in PubMed Google Scholar

Contributions

A.J. conceived and designed the study and the analysis, collected the data, wrote the paper M.V. collected the data and wrote the paper, performed the analysis I.A. collected the data and wrote the paper, performed the analysis V.B. collected the data, designed the figures and formatted the manuscript All authors provided critical feedback and helped shape the research, analysis and manuscript.

Corresponding author

Correspondence to Alen Juginović .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Supplementary information., rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Juginović, A., Vuković, M., Aranza, I. et al. Health impacts of air pollution exposure from 1990 to 2019 in 43 European countries. Sci Rep 11 , 22516 (2021). https://doi.org/10.1038/s41598-021-01802-5

Download citation

Received : 18 May 2021

Accepted : 02 November 2021

Published : 18 November 2021

DOI : https://doi.org/10.1038/s41598-021-01802-5

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Burden of cardiovascular disease attributed to air pollution: a systematic review.

Amir Hossein Khoshakhlagh
Mahdiyeh Mohammadzadeh
Evangelos Oikonomou

Globalization and Health (2024)

Health effects of carbonaceous PM2.5 compounds from residential fuel combustion and road transport in Europe

Jonilda Kushta
Jos Lelieveld

Scientific Reports (2024)

A nationwide epidemiological study on the associations between ambient particulate matter and suicide in Finland

Yoonhee Kim
Harri Antikainen

Nature Mental Health (2024)

Air pollution, cardiorespiratory fitness and biomarkers of oxidative status and inflammation in the 4HAIE study

Lukas Cipryan
Martina Litschmannova
Peter Hofmann

Addressing disparities in the global epidemiology of stroke

Morgan L. Prust
Rachel Forman
Bruce Ovbiagele

Nature Reviews Neurology (2024)

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

Explore articles by subject
Guide to authors
Editorial policies

Information

Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

Active Journals
Find a Journal
Proceedings Series
For Authors
For Reviewers
For Editors
For Librarians
For Publishers
For Societies
For Conference Organizers
Open Access Policy
Institutional Open Access Program
Special Issues Guidelines
Editorial Process
Research and Publication Ethics
Article Processing Charges
Testimonials
Preprints.org
SciProfiles
Encyclopedia

Article Menu

Subscribe SciFeed
Recommended Articles
Google Scholar
on Google Scholar
Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

A visualization approach to air pollution data exploration—a case study of air quality index (pm 2.5 ) in beijing, china.

1. Introduction

2. method and data, 2.1. preliminary analysis, 2.2. hypotheses, 2.3. verification of the hypotheses, 2.4. application layer, 2.5. experimental data, 3. results and discussion, 3.1. general analysis and hypotheses, 3.2. multi-perspective and various visual analysis, 3.2.1. relationship between multiple factors, 3.2.2. temporal characteristics, 3.2.3. spatial characteristics, 4. conclusions, acknowledgments, author contributions, conflicts of interest, appendix a. open source visualization tools, pivottable.js.

Formatting parameters and corresponding return values.
Parameter Interpolation	Return Value
%y: date.getFullYear ()	2013
%m: zeroPad (date.getMonth () +1)	2
%n: mthNames (date.getMonth ())	Feb
%d: zeroPad (date.getDate ())	8
%w: dayNames (date.getDay ())	Fri
%x: date.getDay ()	5 (Sunday is Zero)
%H: zeroPad (date.getHours ())	21 ("PM" is auto-computed)
%M: zeroPad (date.getMinutes ())	10
%S: zeroPad (date.getSeconds ())	30

Click here to enlarge figure

Appendix B. Data quality check of the U-Air data

Appendix c. data processing.

Hypothesis and visualization project design for verification.
Hypotheses	Plots type	Visualization tools	Data Preprocessing
Relationship exists among pollutants and wind speed	Scatter plots	D3.js	By PivotTable.js
There are some regular patterns in time	Heat maps (Circular heat chart, Calendar view)	D3.js
Concentration distributionof pollutants in space	Geovisualization	Openlayers

Mao, F.; Duan, M.; Min, Q.; Gong, W.; Pan, Z.; Liu, G. Investigating the impact of haze on modis cloud detection. J. Geophys. Res. Atmos. 2015 , 120 , 12237–12247. [ Google Scholar ] [ CrossRef ]
Zhang, Y.L.; Cao, F. Fine particulate matter (PM 2.5 ) in China at a city level. Sci. Rep. 2015 , 5 , 14884. [ Google Scholar ] [ CrossRef ] [ PubMed ]
Matejicek, L.; Engst, P.; Jaňour, Z. A GIS-based approach to spatio-temporal analysis of environmental pollution in urban areas: A case study of Prague’s environment extended by LiDAR data. Ecol. Model. 2006 , 199 , 261–277. [ Google Scholar ] [ CrossRef ]
Retalis, A.; Sifakis, N. Urban aerosol mapping over Athens using the differential textural analysis (dta) algorithm on meris-envisat data. ISPRS J. Photogramm. Remote Sens. 2010 , 65 , 17–25. [ Google Scholar ] [ CrossRef ]
Chen, W.; Tang, H.; Zhao, H. Diurnal, weekly and monthly spatial variations of air pollutants and air quality of beijing. Atmos. Environ. 2015 , 119 , 21–34. [ Google Scholar ] [ CrossRef ]
Carslaw, D.C.; Ropkins, K. Openair—An R package for air quality data analysis. Environ. Model. Softw. 2012 , 27–28 , 52–61. [ Google Scholar ] [ CrossRef ]
Janssen, N.A.H.; Fischer, P.; Marra, M.; Ameling, C.; Cassee, F.R. Short-term effects of PM 2.5 , PM 10 and PM 2.5–10 on daily mortality in the Netherlands. Sci. Total Environ. 2013 , 463 , 20–26. [ Google Scholar ] [ CrossRef ] [ PubMed ]
Shi, W.; Wong, M.S.; Wang, J.; Zhao, Y. Analysis of airborne particulate matter (PM 2.5 ) over Hong Kong using remote sensing and GIS. Sensors 2012 , 12 , 6825–6836. [ Google Scholar ] [ CrossRef ] [ PubMed ]
Van Wijk, J.J. The value of visualization. Proc. IEEE 2005 . [ Google Scholar ] [ CrossRef ]
Cai, Z.; Wang, Q.; Weng, M.; Jiang, S.; Du, Q. Information organization and visualization mechanism of electronic map. Geo-Spat. Inf. Sci. 2008 , 11 , 262–268. [ Google Scholar ] [ CrossRef ]
Andrienko, G.; Andrienko, N.; Demsar, U.; Dransch, D.; Dykes, J.; Fabrikant, S.I.; Jern, M.; Kraak, M.J.; Schumann, H.; Tominski, C. Space, time and visual analytics. Int. J. Geogr. Inf. Sci. 2010 , 24 , 1577–1600. [ Google Scholar ] [ CrossRef ]
Kraak, M.J. Geovisualization illustrated. ISPRS J. Photogramm. Remote Sens. 2003 , 57 , 390–399. [ Google Scholar ] [ CrossRef ]
Keim, D.; Andrienko, G.; Fekete, J.-D.; Görg, C.; Kohlhammer, J.; Melançon, G. Visual analytics: Definition, process, and challenges. In Information Visualization ; Springer: Berlin/Heidelberg, Germany, 2008. [ Google Scholar ]
Li, X.; Coltekin, A.; Kraak, M.-J. Visual exploration of eye movement data using the space-time-cube. In Geographic Information Science ; Fabrikant, S.I., Reichenbacher, T., van Kreveld, M., Schlieder, C., Eds.; Springer: Heidelberg, Germany, 2010; Volume 6292, pp. 295–309. [ Google Scholar ]
Zirui, L.; Bo, H.; Dongsheng, J.; Yonghong, W.; Mingxing, W.; Yuesi, W. Diurnal and seasonal variation of the PM 2.5 apparent particle density in Beijing, China. Atmos. Environ. 2015 , 120 , 328–338. [ Google Scholar ]
Pope, C.A.; Burnett, R.T.; Thun, M.J.; Calle, E.E.; Krewski, D.; Ito, K.; Thurston, G.D. Lung cancer, cardiopulmonary mortality, and long-term exposure to fine particulate air pollution. Jama-J. Am. Med. Assoc. 2002 , 287 , 1132–1141. [ Google Scholar ] [ CrossRef ]
Revised Air Quality Standards for Particle Pollution and Updates to the Air Quality Index (AQI). Available online: http://www.epa.gov/airquality/particlepollution/2012/decfsstandards.pdf (accessed on 25 February 2016).
Technical Regulation on Ambient Air Quality Index (on Trial). Available online: http://kjs.mep.gov.cn/hjbhbz/bzwb/dqhjbh/jcgfffbz/201203/W020120410332725219541.pdf (accessed on 25 February 2016).
Jeon, S.J.; Meuzelaar, H.L.C.; Sheya, S.A.N.; Lighty, J.S.; Jarman, W.M.; Kasteler, C.; Sarofim, A.F.; Simoneit, B.R.T. Exploratory studies of PM 10 receptor and source profiling by GC/MS and principal component analysis of temporally and spatially resolved ambient samples. J. Air Waste Manag. Assoc. 2001 , 51 , 766–784. [ Google Scholar ] [ CrossRef ] [ PubMed ]
Chu, D.A.; Kaufman, Y.J.; Zibordi, G.; Chern, J.D.; Mao, J.; Li, C.C.; Holben, B.N. Global monitoring of air pollution over land from the earth observing system-terra moderate resolution imaging spectroradiometer (MODIS). J. Geophys. Res.-Atmos. 2003 , 108 . [ Google Scholar ] [ CrossRef ]
Wang, J.; Christopher, S.A. Intercomparison between satellite-derived aerosol optical thickness and PM 2.5 mass: Implications for air quality studies. Geophys. Res. Lett. 2003 , 30 . [ Google Scholar ] [ CrossRef ]
Othman, N.; MatJafri, M.Z.; Lim, H.S.; Abdullah, K. Retrieval of aerosol optical thickness (AOT) and its relationship to air pollution particulate matter (PM 10 ). In Proceedings of the Sixth International Conference on Computer Graphics, Imaging and Visualization, 2009 CGIV ‘09, Tianjin, China, 11–14 August 2009; pp. 516–519.
Zheng, Y.; Liu, F.; Hsieh, H.-P. U-air: When urban air quality inference meets big data. In 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining ; ACM: Chicago, IL, USA,, 2013; pp. 1436–1444. [ Google Scholar ]
Bostock, M.; Ogievetsky, V.; Heer, J. D-3: Data-driven documents. IEEE Trans. Vis. Comput. Graph. 2011 , 17 , 2301–2309. [ Google Scholar ] [ CrossRef ] [ PubMed ]
Matejicek, L. Spatio-temporal analysis of environmental pollution in urban areas: A case study of the environment in the city of Prague. In Proceedings of the 19th International Congress on Modelling and Simulation (Modsim2011), Perth, Australia, 12–16 December 2011; pp. 1909–1915.
Guo, D.S.; Chen, J.; MacEachren, A.M.; Liao, K. A visualization system for space-time and multivariate patterns (Vis-Stamp). IEEE Trans. Vis. Comput. Graphics 2006 , 12 , 1461–1474. [ Google Scholar ]
Andrienko, G.; Andrienko, N. A general framework for using aggregation in visual exploration of movement data. Cartogr. J. 2010 , 47 , 22–40. [ Google Scholar ] [ CrossRef ]
Kveladze, I.; Kraak, M.J.; van Elzakker, C.P.J.M. A methodological framework for researching the usability of the space-time cube. Cartogr. J. 2013 , 50 , 201–210. [ Google Scholar ] [ CrossRef ]
Li, H.; Fan, H.; Wu, H.; Feng, H.; Li, P. Resdap: A real-time data provision system architecture for sensor webs. In Web and Wireless Geographical Information Systems ; Springer: Heidelberg, Germany; pp. 85–99.
Fan, H.; Li, H. An on-Demand provision model for geospatial multisource information with active self-adaption services. Proc. SPIE 2015 , 9815 . [ Google Scholar ] [ CrossRef ]
Fan, M.; Fan, H.; Chen, N.; Chen, Z.; Du, W. Active on-demand service method based on event-driven architecture for geospatial data retrieval. Comput. Geosci. 2013 , 56 , 1–11. [ Google Scholar ] [ CrossRef ]
Li, L.; Losser, T.; Yorke, C.; Piltner, R. Fast inverse distance weighting-based spatiotemporal interpolation: A web-based application of interpolating daily fine particulate matter PM 2.5 in the contiguous U.S. Using parallel programming and k-d tree. Int. J. Environ. Res. Public Health 2014 , 11 , 9101–9141. [ Google Scholar ] [ CrossRef ] [ PubMed ]
Van Wijk, J.J.; Van Selow, E.R. Cluster and calendar based visualization of time series data. IEEE Symp. Inf. Vis. 1999 , 140 . [ Google Scholar ] [ CrossRef ]
Department of State Air Quality Monitoring Program. Available online: http://www.stateair.net/web/mission/1/ (accessed on 10 December 2015).
Zheng, Y. Urban Air. Microsoft. Available online: http://research.microsoft.com/en-us/projects/urbanair/default.aspx (accessed on 15 October 2015).
World Weather. Available online: http://en.tutiempo.net/ (accessed on 20 January 2016).
Qiao, Z.; Tian, G.; Xiao, L. Diurnal and seasonal impacts of urbanization on the urban thermal environment: A case study of Beijing using MODIS data. ISPRS J. Photogramm. Remote Sens. 2013 , 85 , 93–101. [ Google Scholar ] [ CrossRef ]
Gong, W.; Zhang, T.; Zhu, Z.; Ma, Y.; Ma, X.; Wang, W. Characteristics of PM 1.0 , PM 2.5 , and PM 10 , and their relation to black carbon in Wuhan, central China. Atmosphere 2015 , 6 , 1377–1387. [ Google Scholar ] [ CrossRef ]
Logan, T.; Xi, B.; Dong, X. A comparison of the mineral dust absorptive properties between two Asian dust events. Atmosphere 2013 , 4 , 1–16. [ Google Scholar ] [ CrossRef ]
Cressie, N. The origins of kriging. Math. Geol. 1990 , 22 , 239–252. [ Google Scholar ] [ CrossRef ]
Sanabria, L.; Qin, X.; Li, J.; Cechet, R.; Lucas, C. Spatial interpolation of McArthur’s forest fire danger index across Australia: Observational study. Environ. Model. Softw. 2013 , 50 , 37–50. [ Google Scholar ] [ CrossRef ]
Clark, I. Practical Geostatistics ; Applied Science Publishers: London, UK, 1979; Volume 3. [ Google Scholar ]
Huang, Y.; Nian, P.; Zhang, W. The prediction of interregional land use differences in Beijing: A markov model. Environ. Earth Sci. 2015 , 73 , 4077–4090. [ Google Scholar ] [ CrossRef ]
Wang, W.; Gong, W.; Mao, F.; Zhang, J. Long-term measurement for low-tropospheric water vapor and aerosol by Raman LiDAR in Wuhan. Atmosphere 2015 , 6 , 521–533. [ Google Scholar ] [ CrossRef ]
PivotTable.js. Available online: https://github.com/nicolaskruchten/pivottable (accessed on 24 October 2015).
D3.js. Available online: http://d3js.org/ (accessed on 14 November 2015).

Visualized design for discovery of defects in the U-Air data.
Stations (ID)	Months	Mean AQI of PM
Part (01–22)	All months	Relationship with Month, Day, Hour
All (01–36)	Part (1, 2, 11, 12)	Relationship with Geo-location

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons by Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).

Share and Cite

Li, H.; Fan, H.; Mao, F. A Visualization Approach to Air Pollution Data Exploration—A Case Study of Air Quality Index (PM 2.5 ) in Beijing, China. Atmosphere 2016 , 7 , 35. https://doi.org/10.3390/atmos7030035

Li H, Fan H, Mao F. A Visualization Approach to Air Pollution Data Exploration—A Case Study of Air Quality Index (PM 2.5 ) in Beijing, China. Atmosphere . 2016; 7(3):35. https://doi.org/10.3390/atmos7030035

Li, Huan, Hong Fan, and Feiyue Mao. 2016. "A Visualization Approach to Air Pollution Data Exploration—A Case Study of Air Quality Index (PM 2.5 ) in Beijing, China" Atmosphere 7, no. 3: 35. https://doi.org/10.3390/atmos7030035

Article Metrics

Article access statistics, further information, mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals

An official website of the United States government

Here’s how you know

Official websites use .gov A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS A lock ( Lock A locked padlock ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

JavaScript appears to be disabled on this computer. Please click here to see any active alerts .

Air Quality Trends - How to Interpret the Graphs

The blue band shows the distribution of air pollution levels among the trend sites, displaying the middle 80 percent. The black line represents the average among all the trend sites. Ninety percent of sites have concentrations below the top line, while ten percent of sites have concentrations below the bottom line.

Example of long-term ambient air trends graphic

For each pollutant, the trend statistic is directly related to the level and averaging time of the National Ambient Air Quality Standard (NAAQS). For more information regarding the levels and averaging times for the NAAQS statistics, see the NAAQS Table .

Air Trends Home
Air Quality - National Summary
Air Quality - Cities and Counties
Air Quality Design Values
Previous Air Quality Trends Reports
Carbon Monoxide Trends
Lead Trends
Nitrogen Dioxide Trends
Ozone Trends
Particulate Matter (PM10) Trends
Particulate Matter (PM2.5) Trends
Sulfur Dioxide Trends

Visualizing Air Quality Data - Know through Graph of Air Pollution

Air quality and health are intricately linked. Understanding the real-time status of urban and industrial air quality – and communicating this well – is important for ambient air quality control.

Air quality visualization – the graphical display of data – helps us understand the distribution of air pollutants in the atmosphere. This is hard to do just by looking at a modern air monitor equipment with its digital display.

By combining real-time monitoring data with python programming, one can easily visualize air monitoring data. Interactive graphs can be created which makes it easier to check air quality, and increasingly diverse colors can visually highlight the air quality level. Visualization of data has a resilient expression (more images and more insightful) than the original data table, which is favorable for further analysis of data.

In this article, I will share examples of data visualization that have helped Devic Earth to convey more clearly the message on air quality.

The first example is a heatmap – a plot comparing the data of particulate matter levels (PM2.5) during, the last week of March in 2019 and 2020 - Also here is the judgment of cities with best air quality in india during this period. The purpose of the visualization shown here is to compare the air quality during the lockdown with the same time period in the previous year.

Particulate matter less than 2.5 microns. This pollutant causes cardio-respiratory diseases and is emitted by vehicular pollution and combustion activities

In this graph of air pollution, I have plotted PM2.5 levels in 5 cities during March 22-31 2019 and 2020 using python programming and plotly package.

Figure 1. Heatmap of PM2.5 values during the last week of March in 2019 and 2020 across 5 metropolitan cities. Data source: CPCB monitors

Observation:

In Figure 1, the legend on the right side indicates safe levels of PM2.5 through the color-coding – green indicating safe levels, and purple and brown indicating unsafe levels.

At first glance, the graph of air pollution makes it clear that PM2.5 in 2019 was higher than PM2.5 in 2020 in those 5 cities for the same week in March. For example, the 2019 graph shows PM2.5 levels in Delhi are consistently above 60 µg/m3 and even crosses 160 µg/m3 for a few days, whereas in 2020, the PM2.5 levels have been in the safe zone for most of the days. The CPCB ambient air quality standard in India for 24-hour average PM2.5 is less than 60 µg/m3.

What did I do:

Getting CPCB air quality data for each city for the year 2019 and 2020: CPCB air quality data is easily downloaded from the CPCB website (https://app.cpcbccr.com/ccr/#/caaqm-dashboard-all/caaqm-landing)

Data cleaning and processing it into data frames: Here the missing values and outliers are deleted from the dataset and the date & time values which are in one column are separated into 2 different columns

Creating two subplots and combining them into a single plot in the figures: Two subplots are created for each year setting the x-axis to dates, y-axis to cities, and z-axis i.e. the color to PM2.5 values.

Setting up the layout: This involves setting the color scale to that of Air Quality Index, formatting the text for both the axis in each plot to make it readable in the layouts

Plotting the figures along with the layout : Taking the combined subplots and setting it to the layout created in Step 4, plot the entire graph to get the above visual.

Scatter plot

This is the second graph of air pollution where I have plotted PM2.5 reduction levels against the number of days that Pure Skies has been installed at different industry sectors. Using NumPy, I defined a function to calculate the trend line using OLSM (Ordinary Least Square Method) to show the trend of PM2.5 reductions over time.

PM2.5 reductions improve over time, especially in the heavy machine industries where the baseline graph of air pollution is higher. The trend line which shows a steady increase also indicates the same. The trend line indicates the PM2.5 reductions of 40-60% in the first 150 days. After that, there is an 80% reduction in PM2.5.

Divide the dataset based on the type of industries.

Define a function to calculate the trendline (y = mx + c, where m is the slope and c is the intercept) using Ordinary Least Square Methods (source:https://docs.scipy.org/doc/numpy/reference/generated/numpy.linalg.lstsq.html)

For the plot, using each divided dataset as a separate trace for scatter plot, with the x-axis being the number of days and the y-axis being the PM2.5 reduction, combine all the traces as one array for the figure.

Set the format for the axis and texts in the layout to make it readable.

Run the function created in step 2 giving the values of x and y in the graph, then add a trace for line plot keeping the x-axis the same and y as m*x + c.

Plot the entire figure with the layout.

“A picture speaks a thousand words”

Data visualizations make big and small data simpler for the human brain to comprehend. Visualization also makes it easier to identify patterns, trends, and outliers in sets of data. Good data visualizations should place meaning into complex datasets so that their significance is well-defined and brief.

In terms of business, data visualization speeds up the process of decision making and analytics where instead of calculating for trends or patterns one can easily see it on a graph, saving a measurable amount of time. Since visualization is something everyone can understand it enhances communication across departments in the business, thus making it more efficient.

In terms of air quality – which often cannot be seen – data visualization eliminates that disadvantage, hence displaying the quality of air a person is breathing. It even displays the efficacy of air control equipment, which will help businesses when it comes to investing in air control equipment.

Looking for a Devic Earth recommended and internally used Indoor Air Quality Monitor?

Check this out and book a free consultation with us!

(This article is written by Vidhya Sreenivasan with updates from Shashank Aggarwal)

Articles you might like:

5 policy recommendations for better air quality post-COVID

Work in a factory? Understand how Pure Skies can help you

How can clean air help you cut costs after the lockdown

Author- Vidhya Sreenivasan

Mining Processes: Ambient Air Pollution, Emissions, and the Solution

Post Comments

Air Pollution and Child Mental Health

Vidhya Sreenivasan Technical Analyst

Pure Skies Technology Crushed Air Pollution by 81% at A Heavy Manufacturing Plant!

Delhi reels under twin emergencies: Air pollution & COVID-19

Air pollution is a critical risk factor for COVID-19 infections and mortality, according to...

Our Social Channel

Subscribe for our newsletter.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Publications
Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

Advanced Search
Journal List
Int J Environ Res Public Health

Visualization and Analysis of Air Pollution and Human Health Based on Cluster Analysis: A Bibliometric Review from 2001 to 2021

1 Zhou Enlai School of Government, Nankai University, Tianjin 300071, China

2 College of Management and Economy, Tianjin University, Tianjin 300072, China

Kevin Huang

3 School of Accounting, Economics and Finance, University of Wollongong, Sydney, NSW 2522, Australia

4 School of Marxism, Hangzhou Medical College, Hangzhou 310053, China

Tiantong Xu

5 School of E-Business and Logistics, Beijing Technology and Business University, Beijing 100048, China

Zhenni Chen

6 School of Economics and Finance, Xi’an Jiaotong University, Xi’an 710061, China

7 School of Economics and Management, Beijing Institute of Petrochemical Technology, Beijing 102617, China

Associated Data

The original contributions presented in the study are included in the supplementary material . Further inquiries can be directed to the corresponding author.

Bibliometric techniques and social network analysis are employed in this study to evaluate 14,955 papers on air pollution and health that were published from 2001 to 2021. To track the research hotspots, the principle of machine learning is applied in this study to divide 10,212 records of keywords into 96 clusters through OmniViz software. Our findings highlight strong research interests and the practical need to control air pollution to improve human health, as evidenced by an annual growth rate of over 15.8% in the related publications. The cluster analysis showed that clusters C22 (exposure, model, mortality) and C8 (health, environment, risk) are the most popular topics in this field of research. Furthermore, we develop co-occurrence networks based on the cluster analysis results in which a more specific keyword classification was obtained. These key areas include: “Air pollutant source”, “Exposure-Response relationship”, “Public & Occupational Health”, and so on. Future research hotspots are analyzed through characteristics of the cluster groups, including the advancement of health risk assessment techniques, an interdisciplinary approach to quantifying human exposure to air pollution, and strategies in health risk assessment.

1. Introduction

The negative effects of air pollution on human health are well documented [ 1 , 2 ]. Harmful air pollutants, such as PM 2.5 , PM 10 , SO 2, and NO X , escaping into the environment through natural and human activities may adversely affect human health [ 3 ]. Air pollution has acute and chronic effects on various systems and organs. These range from upper respiratory tract irritation to chronic respiratory and heart diseases [ 4 ], lung cancer [ 5 ], childhood acute respiratory infections [ 6 ] and adult chronic bronchitis, exacerbating existing cardiopulmonary diseases or asthma attacks [ 7 ], and even causing serious mental illness [ 8 ]. According to the official statistics of the World Health Organization, the number of people killed by air pollution is as high as 7 million every year, and 9 out of every 10 people in the world still breathe air containing high levels of pollutant concentration [ 9 ].

Due to the growing public concern about the adverse health consequences caused by air pollution, an increasing number of publications have examined the correlation between exposure to air pollutants and the incidence and mortality of various diseases [ 10 , 11 , 12 , 13 , 14 , 15 , 16 ]. For example, some studies have shown that exposure to particulate matter and ozone in the air is associated with increased mortality and hospitalization due to respiratory and cardiovascular disease [ 17 , 18 ]. Particulate matter is likely to penetrate the lungs and cardiovascular system, leading to diseases such as stroke, heart disease, lung cancer, and chronic obstructive pulmonary disease [ 19 , 20 ]. Nevertheless, most of the publications focusing on specific fields are scattered, and the interrelationship between them is not yet clear. Although many experts have summarized the results of research in this field [ 21 , 22 , 23 , 24 ], there are still few studies on the application of bibliometric methods for analysis. So far, there are only two articles that review air pollution and human health with a bibliometric method. Han et al. [ 25 ] used bibliometric methods to analyze the relevant literature in the field of health effects caused by PM 2.5 since 2000. Dhital et al. [ 26 ] carried out a bibliometric analysis of 2179 documents published during the last two decades. The limitation of the above two studies is that they mainly focus on the performance of literature from the perspectives of international collaboration, journal distribution, authorship cooperation, etc., which belong to the basic descriptive results. Therefore, in this paper, to further examine the hot research topics, we combine machine learning with bibliometric methods to quantitatively review the literature in this field. The quantitative methods are more conducive to helping us better understand the air pollution and health research landscape.

This paper explores the development of this field through statistical methods from a macro perspective. The primary objective of this study is to give a comprehensive overview of the research publications on air pollution and human health published during the last two decades (2001–2021). The main contributions of this paper are as follows: (1) This study aims to provide a qualitative and quantitative evaluation of the current research progress and trends in air pollution and health research. (2) Some publication characteristics, including countries, funding agencies, journals, and international collaborations, are presented. (3) The study applies the principle of machine learning to divide 10,212 records into 96 clusters and identifies the positional relationship between them. (4) Based on the clustering results, we develop co-occurrence networks and obtain more specific keyword classification results by keyword frequency, relationship, and semantic analysis to obtain future research hotspots.

2. Materials and Methods

2.1. data collection.

The Web of Science Core Collection provides a variety of records for each publication, including author information, journals, citations, and institutional affiliations. We mainly search for articles from SCI and SSCI, published in English from 2001 to 2021. The study used keywords (i.e., “air pollution*” or PM 2.5 or ozone or “particular matter”) and (health or mortality or fatality or death or epidemiology or fitness or morbidity) to search and collect research articles. A total of 46,934 pieces of literature were identified through database searches. After excluding 11,963 pieces of literature of the non-article type and an additional 2566 via screening of the titles and abstracts, finally, 14,955 articles were full-text screened for eligibility (See Figure 1 , Figures S1 and Table S1 ).

An external file that holds a picture, illustration, etc.
Object name is ijerph-19-12723-g001.jpg

Framework of literature search, analysis, and interpretation.

In addition, OmniViz (BioWisdom Ltd., Cambridge, UK) was used to extract and cluster keywords from the articles. OmniViz is an advanced visual informatics software package that is designed to provide visualization of digital data, categorical data, genomic sequences, chemical structures, and text documents [ 27 ]. It can analyze large data sources through different clustering methods. We used OmniViz to identify important topics and hot research areas. In addition, the clustering method was used to measure the similarity of two records in a high-dimensional space. To achieve data visualization, we used Galaxy and Thememap. Galaxy provides relationships between lots of records, and Thememap identifies the most important topics in the field. Please see Figure S2 in the supplementary file for details on methodology and software.

2.2. Impact Factors

The impact factor (IF) and h-index are well-recognized indicators that are closely related to the bibliometric analysis. The IF is a useful indicator to quantify the rank and quality of a journal. The IF of a journal is calculated by dividing the citation count of the current year by the number of published articles in the journal during the previous two years [ 28 ]. It is created by the Institute of Scientific Information (ISI). A higher IF usually reflects a journal’s higher quality in various research fields. The h-index means that ‘h’ of one’s total articles are cited at least ‘h’ times. It is a popular indicator to measure the performance of a scientist and has been widely used to evaluate the academic performance of a journal or a country.

2.3. Social Network Analysis (SNA)

Social network analysis (SNA) refers to a computable analysis method based on multidisciplinary fusion theories and methods to understand the formation of various human social relationships, behavioral characteristics analysis, and the laws of information transmission. It aims to quantify the network’s structure features and the dynamic interactions among network vertices. Due to the development of social network theory, SNA has been widely used to analyze academic collaboration in different fields [ 29 ]. By using a variety of measurement metrics, the contributions from different countries, institutions, and scientists can be evaluated.

In this study, SNA is used to evaluate collaboration among different countries and institutions [ 30 ], which includes two steps. The first step is information extraction. The country and institutional information for each author was extracted using BibExcel so that the visualization effect of academic cooperation among different countries can be presented. The second step is to draw a cooperation diagram with the input data from BibExcel using Pajek to visualize their cooperation patterns.

3. Results and Discussions

3.1. the performance of related publications.

In recent years, the issues of air pollution have gradually drawn public attention. Our study focuses on the effect of air pollution on health from the perspective of bibliometrics so as to understand the current research status and future research trends. Figure 2 shows the annual number (NO) of articles published between 2001 and 2021, the total number of citations (TC) for the articles, and the average citation count (ACPP) for each article. It can be observed that the NO has grown slowly in the first 12 years and has increased rapidly with a growth rate of more than 9% since 2008. The number of articles published after 2009 accounted for nearly 85% of the total number of published articles. In addition, the TC grew steadily during the first 11 years, peaked in 2008 and 2018, and then gradually declined. Due to the increasing number of published articles, the ACPP showed a downward trend as a whole.

An external file that holds a picture, illustration, etc.
Object name is ijerph-19-12723-g002.jpg

Numbers of the NO, TC, and ACPP during the period of 2001–2021. Notes: NO represents the number of published articles, TC represents the total citations of articles, and ACPP represents the average citation counts of each article.

3.2. Publication Features of Different Countries

The number of publications reflects the academic strengths and attentions of each country in the field. The top five most productive countries are the United States (6932 publications), China (4350 publications), the United Kingdom (2206 publications), Canada (1035 publications), and Italy (974 publications). These top five countries published a total of over 10,000 articles, accounting for 78.55% of all publications. Among these productive countries, the United States outperforms others in the total number of published articles on air pollution during 2001–2021 [ 31 ]. It can be observed that China’s publications have grown rapidly since 2014, with an average growth rate of over 30%. Moreover, the amount of literature published by Chinese scholars gradually approached the USA’s productivity in 2021.

We applied SNA to analyze the international collaboration among the 20 most productive countries during the period 2001–2021 ( Figure 3 ). The lines connecting the countries represent their cooperation, and the line thickness indicates the degree of collaboration [ 32 , 33 ]. Collaboration was determined by the affiliations of the co-authors, and all countries or institutes stand to benefit if one publication is a collaborative study [ 34 , 35 ]. These 20 productive countries worked closely with each other, particularly the U.S.A., China, Canada, Germany, and the U.K. The U.S.A. was the center of this collaboration network and the leader of air pollution research in cooperation with the other productive countries. The U.S.A. and China had the closest collaboration, followed by the U.S.A.–U.K., U.S.A.–Canada, and the U.S.A.–Germany.

An external file that holds a picture, illustration, etc.
Object name is ijerph-19-12723-g003.jpg

International collaboration according to social network analysis.

3.3. The Performances of Different Journals

The top 20 most productive journals are shown in Table 1 . These productive journals account for 63.4% of the total related publications. In particular, Atmospheric Environment is the most productive journal with a count of 1506 (10.07%) articles. Other dominating journals include Environmental Health Perspectives , Science of the Total Environment , and Environmental Research . The IF of the journal is not the only index to reflect the journal’s influence in the field. Therefore, we calculated the average citations to better reflect the journal’s influence. The results showed that Epidemiology and Environmental Health Perspectives have the highest average citations (86.25 and 81.30), followed by the American Journal of Epidemiology , Occupational and Environmental Medicine , and Environmental Science and Technology .

The top 20 most productive journals.

Journal Titles	Records	Percentage	IF2022	Average Citations
International Journal of Environmental Research and Human Health	1506	10.07%	4.61	15.70
Science of the Total Environment	1415	9.46%	10.75	26.33
Atmospheric Environment	1240	8.29%	5.75	35.20
Environmental Research	1110	7.42%	8.43	28.85
Environmental Health Perspectives	971	6.49%	11.03	81.30
Environment International	861	5.75%	9.62	23.44
Environmental Pollution	731	4.88%	9.98	25.22
Environmental Science and Technology	632	4.22%	11.35	46.15
Environmental Science and Pollution Research	538	3.59%	5.19	11.72
Journal of the Air and Waste Management Association	510	3.41%	2.63	35.67
Environmental Science and Pollution Research	486	3.24%	5.19	12.56
Environmental Health	435	2.90%	7.12	20.90
Atmospheric Chemistry and Physics	407	2.72%	7.19	31.64
Air Quality Atmosphere and Health	359	2.40%	5.80	13.68
Journal of Exposure Science and Environmental Epidemiology	336	2.24%	6.37	22.73
Epidemiology	321	2.15%	4.99	86.25
Journal of Toxicology and Environmental Health-Part A-Current Issues	278	1.85%	2.71	24.15
Aerosol and Air Quality Research	266	1.77%	2.59	10.33
Environmental Monitoring and Assessment	227	1.52%	1.80	13.57
Occupational and Environmental Medicine	202	1.35%	3.97	50.23
American Journal of Epidemiology	198	1.32%	4.32	67.28

1 Note: IF2022 represents the published impact factor for 2022. The impact factor is a measure of the importance of a journal. The impact factor is calculated by dividing the number of times the articles are cited in the last two years by the total number of publications in those two years.

Furthermore, the field of air pollution and health is typically an interdisciplinary area. According to the statistics ( Figure 4 ), the largest proportion of research areas are in Environmental Sciences and Ecology; Public, Environmental, and Occupational Health; Meteorology and Atmospheric Sciences; and Toxicology, which account for 93.2% of the total number of publications. Among them, the Environmental Sciences and Ecology research area maintains an average growth rate of more than 15%, occupying the most important position.

An external file that holds a picture, illustration, etc.
Object name is ijerph-19-12723-g004.jpg

The top 5 research areas of publications.

3.4. Institutions’ Performances

The performances of the top 20 most productive institutions are listed in Table 2 . Most institutions are from the productive countries shown in the previous section. Among them, twelve institutions are located in the U.S.A and three are from China. Harvard University is the most productive research organization with 1263 publications, followed by the University of California System, the United States Environmental Protection Agency, and the Chinese Academy of Sciences. In the U.S.A., universities and government research institutes, such as the United States Environmental Protection Agency, are the main forces in the field. Additionally, several European countries, such as England, the Netherlands, and Switzerland, have mature experience in air pollution prevention and reducing its negative impact on health. Therefore, it is not surprising to see that 4 institutions in European countries are listed among the top 20 most productive institutions.

The top 20 most productive institutions.

Rank	Institutions	Country	Records
1	Harvard University	USA	1263
2	University of California System	USA	1230
3	United States Environmental Protection Agency	USA	1205
4	Chinese Academy of Sciences	China	1183
5	University of California Berkeley	USA	1005
6	University of Northern California	USA	956
7	University of London	UK	901
8	Peking University	China	874
9	Helmholtz Association	Germany	856
10	University of Washington	USA	764
11	University of Washington—Seattle	USA	732
12	Utrecht University	The Netherlands	724
13	Health Canada	Canada	683
14	University of North Carolina at Chapel Hill	USA	664
15	University of Southern California	USA	628
16	Imperial College London	UK	540
17	Emory University	USA	529
18	Johns Hopkins University	USA	487
19	Fudan University	China	430
20	Columbia University	USA	425

3.5. Research Hotspots Analysis

3.5.1. keyword clustering.

A machine learning-based cluster analysis was carried out on the keywords of 14,955 research articles using software named OmniViz. A total of 10,212 records were obtained and divided into 96 clusters. The article selected clusters with more than 50 records, for a total of 29 clusters, as shown in Table 3 . Moreover, we also give top 20 frequent keywords list ( Table S2 ). The cluster C22 (exposure, model, mortality) has the most publications with a total of 3629, which accounts for 37.1% of the total. In addition, C8 (health, environment, risk), C92 (health, environment, risk), C35 (exposure, person, particle), C6 (model, ozone, emit), and C64 (aerosol, source, dust) are recorded more than 300 times. These clusters were also usually selected as research topics. In addition, the distance of each cluster in the galaxy map can reflect their correlations. If their locations are closely related to their research relevance, their relevance is very high. On the contrary, these research themes are not strongly related. The clusters of publications recorded with more than 50 instances are shown in Figure 5 and marked as yellow. It can be concluded that C8, C47, C81, and C25 are closed to each other and concentrated at the top of the galaxy map. This indicates that keywords such as “environment, health, and risk” are often associated with “exposure, chronic, disease, heart,” and so on. Similarly, located in the middle and lower parts of the galaxy map, C89, C64, C35, C58, and C56 are also closely linked. This suggests that the relationships between the keywords of “particle, ozone, dust”, and other related pollutants and “aerosol, concentration, source” are very close. In addition, clusters C8 and C92 show the same term labels, both of which are “health, environment, risk”. The number of records is above 500. However, they are from different collections of publications at different positions on the galaxy map ( Figure 5 ). Each pollutant has different sources and measurement methods at the pollutant level [ 36 ]. The health level involves various diseases, and the related model methods are often employed in related research [ 37 ].

An external file that holds a picture, illustration, etc.
Object name is ijerph-19-12723-g005.jpg

Keywords Galaxy map (records > 50).

The selected clusters (records > 50) by OmniViz analysis.

No.	Records	Major Terms	No.	Records	Major Terms
C6	456	Model, ozone, emit	C56	242	Ozone, water, surface
C8	722	Health, environment, risk	C58	181	Particle, source, aerosol
C17	76	Carbon, organic carbon, aerosol	C59	93	Model, exposure, process
C20	87	Standard, management, ventilation	C64	336	Aerosol, source, dust
C22	3629	Exposure, model, mortality	C69	162	Risk, cancer, dust
C23	107	Mortality, heat, weather	C74	157	Environment, social, health
C24	69	Aerosol, concentration, process	C77	89	Fuel, energy, household
C25	52	Climate, disease, global	C81	136	Disease, heart, chronic
C32	121	Mortality, ozone, risk	C84	182	Sense, aerosol, source
C35	482	Exposure, person, particle	C85	97	Emit, vehicle, ozone
C37	68	Emit, vehicle, industry	C87	230	Species, atmospheric, carbon
C43	115	Heat, wave, mortality	C89	172	Particle, aerosol, concentration
C47	54	Health, exposure, environment	C92	537	Health, environment, risk
C48	121	Environment, exposure, particle	C95	56	Dioxide, ozone, carbon
C50	58	Smoke, tobacco, chronic obstructive pulmonary disease

In addition, as shown in Figure 6 , a theme map can identify the main topics in the research field of air pollution and health, which is an effective complement to the Galaxy visualization. The height of the peak depends on the intensity of the topic and the concentration of information at that location. It can be observed that the four highest peaks are: “health, environment, risk”, “exposure, model, mortality”, “exposure, ozone, person”, and “model, ozone, emit”. The results are very similar to the cluster analysis; however, there are some differences in the height order. The main reason is that, although C22 has the most records, it is not closely related to the surrounding clusters. The cluster C8 is more intensive at this position, resulting in a stronger theme, which led to the highest peak. Compared with others, the peaks around “exposure, person, particle” are significantly denser. There are more valleys surrounding the rest of the peaks, which suggests that such research is highly relevant and often involves an interdisciplinary approach.

An external file that holds a picture, illustration, etc.
Object name is ijerph-19-12723-g006.jpg

Theme map of keywords.

3.5.2. Relationship of Keywords among Different Groups

Through keyword clustering, 10,212 records were divided into 96 clusters. In the keywords Galaxy map ( Figure 7 ), we can roughly divide them into three groups, i.e., Group I, Group II, and Group III, based on the positional relationships of different clusters. Although the clusters located at the lower left of the Galaxy map are very dense, most of these clusters appear less than 50 times and, therefore, will not be analyzed further. Within each group, we develop co-occurrence networks and obtain more specific keyword classifications based on keyword frequency, relationship, and semantic analysis. The research hotspots are analyzed through the characteristics of three groups.

An external file that holds a picture, illustration, etc.
Object name is ijerph-19-12723-g007.jpg

Group I research topics.

(1) Group I

The impact of the deterioration of the ecological environment, especially of the air quality, on human health has received a growing level of attention. The damage to health caused by air pollution further increases the degree of health inequalities among groups of different income levels [ 38 , 39 ]. As shown in Figure 7 , “environment and health” are recognized as the central keywords of Group I because of their high frequencies and close relationships with other research topics. Focusing on two central keywords, we identified four relevant research areas, i.e., “Air pollutant source”, “Exposure–Response relationship”, “Health & Mortality”, and “Cost & Benefit”. In terms of air pollutant sources, outdoor sources often refer to the cluster C37 (Emit, vehicle, industry), mainly including industry and vehicle emissions. Household pollution sources, such as the keyword clusters C77 (Fuel, energy, household) and C20 (Standard, management, ventilation), are often combined with specific issues, such as fuel burning, use of building materials, chemicals, and ventilation in household activities [ 40 ]. Most air pollutants monitored by remote sensing technology are always assessed by the exposure-response function [ 41 ], such as the clusters C6 (Model, Ozone, emit) and C89 (Particle, aerosol, concentration), to reflect such research trends. In addition, some studies have stated that air pollution can cause increased morbidity, including heart disease and chronic diseases [ 42 ]. And in Table S3 , we also summarized highly co-cited documents among research clusters in Clusters I.

(2) Group II

As shown in Figure 8 , “exposure” is recognized as a core keyword, and three keywords with high frequency are closely related (i.e., particle, aerosol, source). With these four keywords as the core, four relevant research areas have been formed, namely “Air pollution source”, “Air pollution monitoring”, “Particulate matter concentration”, and “Atmospheric aerosol”. The current methods of air pollution control focus on source management, so identifying air pollution sources is still an important research area [ 43 ].Moreover, dust in cities can also increase the short-term mortality of vulnerable populations, such as extreme dust episodes in high-density Asian cities [ 44 ]. In order to formulate reasonable air pollution control policies, real-time monitoring of air quality becomes more important. In addition to common air monitoring stations [ 45 ], some studies suggest that air quality monitoring can be performed in innovative ways, such as through social media [ 46 ] and mobile sensors [ 47 ]. Though PM and ozone are the main monitored pollutants, related studies have gradually evolved from single-pollutant to multi-pollutant collaborative studies, such as ozone, particle, carbon monoxide, PM 2.5 , PM 10 , etc. [ 48 ]. As shown in Figure 8 , there is a strong relationship between “particle” and “aerosol”. Studies have found that severe haze pollution incidents were mainly caused by the formation of secondary aerosols. Moreover, in Table S4 , we also summarized highly co-cited documents among research clusters in Clusters II.

An external file that holds a picture, illustration, etc.
Object name is ijerph-19-12723-g008.jpg

Group II research topics.

(3) Group III

As shown in Figure 9 , “aerosol” is the central word in this section. Around the central keyword, four related research areas can be identified, including “Atmospheric physics”, “Atmospheric chemistry”, “Health & Mortality”, and “Public & Occupational Health”. Atmospheric physics and atmospheric chemistry are the basic disciplines in related research. They can explain the formation and transmission mechanisms of atmospheric pollutants and provide a theoretical basis for controlling air pollution [ 49 ]. The research field “Health & Mortality” was once again emphasized in the clustering results. Compared with Group I, this part has more diseases mentioned. The measurement of health risks from the mortality index [ 50 ] gradually concentrated on the incidence of specific diseases, including cancer, heart disease, respiratory diseases, and so on. Further, based on the complexity of multi-pollutant collaborative research [ 51 ], studies must develop more targeted evaluation models and apply new model fusion methods, i.e., the air pollution mortality/incidence risk (Ri-MAP) model [ 52 ], comprehensive health risk index, exposure–response coefficient [ 53 ], CMAQ/GCAM evaluation model [ 54 ], and so on. We are pleasantly surprised to find that public and occupational health research is further valued in Group III. Some studies currently classify occupational characteristics and social status to assess the health effects of air pollution on different groups [ 55 ]. For example, the socioeconomic status of parents, including education, income, and living area, has an impact on children’s health, which suggests that health can play an important role in the intergenerational transmission of economic status [ 56 ]. In addition, the highly co-cited documents among research clusters in Cluster II have been summarized in Table S5 .

An external file that holds a picture, illustration, etc.
Object name is ijerph-19-12723-g009.jpg

Group III research topics.

4. Conclusions

4.1. summary.

This study conducted a bibliometric analysis of 14,955 articles on air pollution and health from 2001 to 2021. In the past two decades, the most productive country, the most productive institution, and the most productive journal are the United States, Harvard University, and Atmospheric Environment , respectively. We used OmniViz to cluster the keywords of 10,212 records, and the results show that the clusters with the most occurrences were “Exposure, model, mortality”, “Health, environment, risk”, and “Model, Ozone, emit”. By observing the Thememap and Galaxy visualization results, additional popular topics in the study include “Health, environment, risk” and “Exposure, person, particle”. Based on the clustering results, we developed co-occurrence networks and obtained more specific keyword classification results according to keyword frequency, relationship, and semantic analysis. We have identified the most influential areas, such as “Air pollutant source”, “Exposure–Response relationship”, “Public & Occupational Health”, and so on. The research hotspots are analyzed through the characteristics of three groups of clusters. Indeed, this paper provides a qualitative and quantitative evaluation of the current research progress and trends in air pollution and health research.

4.2. Limitations and Future Research Directions

Nonetheless, although bibliometric analysis is an effective method for reviewing the literature, it is not without limitations. First, the bibliometric data from the Web of Science (including SCI-EXPAND and SSCI) are not produced exclusively for analysis, thus the data may contain errors, wherein the presence of errors is bound to influence any analysis performed using such data. Thus, to mitigate errors, we have carefully cleaned the bibliometric data that we searched. For example, we remove duplicates and erroneous entries. Second, the nature of the bibliometric method is in itself a limitation. We noticed that the qualitative assertions of bibliometrics can be subjective given that bibliometric analysis is quantitative in nature, whereas the relationship between qualitative and quantitative results is often unclear. To solve these problems, we tried to combine machine learning and the bibliometric method to make a more accurate results analysis, but there is still room for method improvement. Third, bibliometric studies can only offer a certain period forecast of the research field, and thus scholars should avoid making overly ambitious assertions in their research field. Notwithstanding these limitations, the bibliometric method can help us to overcome the fear of large bibliometric datasets and to pursue retrospectives of air pollution and human health. Indeed, the bibliometric methods can not only facilitate knowledge in this field but also help us to better understand the research trend. We take a short yet significant step in that direction.

Based on the bibliometric analysis, three main future directions of air pollution and health are identified in this study. The first future research direction is the advancement of health risk assessment techniques. At present, the air monitoring network does not take into account health factors such as changes in air pollution components, public medical data, and population activity patterns [ 57 ]. There is a lack of dynamic data to support the health risk assessment as the monitoring network of air pollution’s health impact is far from mature [ 58 ]. Health risk assessment techniques currently face the challenge of transformation from qualitative research to quantitative research. Future research can be conducted to examine the route and trajectory of pollutant exposure to determine the actual intake and intake coefficient of air pollutants [ 59 ]. Subsequently, sophisticated time-activity patterns and individual exposure monitoring techniques should be employed more widely to determine accurate exposure doses [ 60 ]. More suitable biological targets should be identified to establish a quantitative relationship between the growth in the concentration of pollutants and the increase in mortality and disease prevalence.

The second future research direction is an interdisciplinary approach to quantifying human exposure to air pollution. Quantifying human exposure to air pollutants is a challenging task [ 61 ]. Exposure results from multifaceted relationships and interactions between environmental and human systems, adding complexity to the assessment process [ 62 ]. For assessment of the health risks of air pollution, related studies might evolve from epidemiology research to toxicology research. From the perspective of environmental toxicology, future studies can be undertaken on the biological mechanism of the bioavailability and toxicity of particulate matter. In order to establish more accurate quantitative models, real-time air pollution health risks can be described using mobile air pollution monitoring techniques, meteorological information, and land use information [ 61 ]. Air quality changes and mobile monitoring allow relevant departments to respond to these changes quickly. Consequently, the threshold for the impact of different pollutants on human health can be determined more accurately.

The third future research direction is management strategies in health risk assessment [ 63 ]. Proven effective management strategies include the establishment and enforcement of air standards, reduction in emissions from coal-fired power plants and other stationary sources, banning of the use of polluting fuels in urban centers, improvements to access to public transportation, and so on [ 64 ]. Future air pollution prevention strategies will emphasize integration and cooperation. Air pollution control in some key areas should pay attention to regional and departmental cooperation. Policy boundaries will become increasingly blurred, and mandatory regulatory policies may also need to incorporate economic incentives. A simple economic incentive policy will have a small audience and need to be combined with other policies to innovate the current policy tool system. Risk management strategies will shift from simple restrictions and prohibitions to more flexible multi-policy coordination.

Acknowledgments

We express our gratitude for the Software support from Green Development Center at Tianjin University.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/ijerph191912723/s1 . Figure S1: Framework of literature search, analysis, and interpretation. Note: Science Citation Index (SCI); Social Science Citation Index (SSCI). Figure S2: Methodology and software support. Table S1: Keywords used in the search and results. Table S2: The trend of 15 most frequent keywords during 2001–2021. Table S3: Highly co-cited documents among research clusters in Cluster I. Table S4: Highly co-cited documents among research clusters in Cluster II. Table S5: Highly co-cited documents among research clusters in Cluster III.

Funding Statement

This study was funded by the Fundamental Research Funds for the Central Universities (Grant No. 63222036), China Postdoctoral Science Foundation (Grant No. 2022M710072; No. 2021M692568; No. 2020M670636), National Social Science Foundation (Grant No. 21CZZ007); and MOE (Ministry of Education in China) Project of Humanities and Social Sciences (No. 21YJC630014). And the APC was funded by China Postdoctoral Science Foundation (Grant No. 2022M710072).

Author Contributions

All authors contributed equally to this work. D.L. was responsible for analyzing the results with machine learning and bibliometric methods. K.C. made a substantial and meaningful contribution to the revision. K.H. was responsible for supervision. H.D. was responsible for reviewing, editing, and visualization with software. T.X. and Z.C. were responsible for searching, collecting, and screening related literature. Y.S. was responsible for writing, editing, and reviewing. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Data availability statement, conflicts of interest.

The authors declare no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

IEEE Account

Change Username/Password
Update Address

Purchase Details

Payment Options
Order History
View Purchased Documents

Profile Information

Communications Preferences
Profession and Education
Technical Interests
US & Canada: +1 800 678 4333
Worldwide: +1 732 981 0060
Contact & Support
About IEEE Xplore
Accessibility
Terms of Use
Nondiscrimination Policy
Privacy & Opting Out of Cookies

A not-for-profit organization, IEEE is the world's largest technical professional organization dedicated to advancing technology for the benefit of humanity. © Copyright 2024 IEEE - All rights reserved. Use of this web site signifies your agreement to the terms and conditions.

Analysis of Air Pollution Data in India between 2015 and 2019

1 Center for Policy Research on Energy and Environment, School of Public and International Affairs, Princeton University, Princeton, NJ 08544, USA 2 Department of Civil and Environmental Engineering, Princeton University, Princeton, NJ 08544, USA

Copyright The Author(s). This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0) , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are cited.

Download: PDF | Supplemental Material

Sharma, D., Mauzerall, D. (2022). Analysis of Air Pollution Data in India between 2015 and 2019. Aerosol Air Qual. Res. 22, 210204. https://doi.org/10.4209/aaqr.210204

Analysis of PM 10 , PM 2.5 , SO 2 , NO 2 and O 3 measurements across India from 2015–2019.
First comprehensive analysis of Indian government and US Air-Now data.
More national ambient air quality standard exceedances in north than south India.
Provides baseline for evaluation of mitigation measures and atmospheric models.

India suffers from among the worst air pollution in the world. In response, a large government effort to increase air quality monitoring is underway. We present the first comprehensive analysis of government air quality observations from 2015–2019 for PM 10 , PM 2.5 , SO 2 , NO 2 and O 3 from the Central Pollution Control Board (CPCB) Continuous Ambient Air Quality Monitoring (CAAQM) network and the manual National Air Quality Monitoring Program (NAMP), as well as PM 2.5 from the US Air-Now network. We address inconsistencies and data gaps in datasets using a rigorous procedure to ensure data representativeness. We find particulate pollution dominates the pollution mix across India with virtually all sites in northern India (divided at 23.5°N) exceeding the annual average PM 10 and PM 2.5 residential national ambient air quality standards (NAAQS) by 150% and 100% respectively, and in southern India exceeding the PM 10 standard by 50% and the PM 2.5 standard by 40%. Annual average SO 2 , NO 2 and MDA8 O 3 generally meet the residential NAAQS across India. Northern India has (~10%–130%) higher concentrations of all pollutants than southern India, with only SO 2 having similar concentrations. Although inter-annual variability exists, we found no significant trend of these pollutants over the five-year period. In the five cities with Air-Now PM 2.5 measurements - Delhi, Kolkata, Mumbai, Hyderabad and Chennai, there is reasonable agreement with CPCB data. The PM 2.5 CPCB CAAQM data compares well with satellite derived annual surface PM 2.5 concentrations (Hammer et al. , 2020), with the exception of the western desert region prior to 2018 when surface measurements exceeded satellite retrievals. Our reanalyzed dataset is useful for evaluation of Indian air quality from satellite data, atmospheric models, and low-cost sensors. Our dataset also provides a baseline to evaluate the future success of National Clean Air Programme as well as aids in assessment of existing and future air pollution mitigation policies.

Keywords: Air pollution, India, surface observations, CPCB, continuous and manual data, US AirNow

1 INTRODUCTION

Concerns over poor air quality in India have increased over the past few years with increasing evidence of the adverse impacts on health (Balakrishnan et al. , 2014; Chowdhury and Dey, 2016; Balakrishnan et al. , 2019), agricultural yields (Avnery et al. , 2011, 2013; Ghude et al. , 2014; Gao et al. , 2020) and the economy (Pandey et al. , 2021). Rapid growth and industrialization in India have resulted in some of the most polluted air in the world. Projections forecast further decreases in air quality and a 24% increase in PM 2.5 associated premature mortalities by 2050 relative to 2015 (GBD MAPS Working Group, 2018; Brauer et al. , 2019). According to recent estimates based on the Global Exposure Mortality Model (GEMM), total premature mortality due to ambient PM 2.5 exposure in India increased approximately 47% between 2000 and 2015 (Chowdhury et al. , 2020). Surface O 3 concentrations are also likely to increase with growing industrial emissions and increasing temperatures due to climate change resulting in additional stress on agricultural yields and public health (Avnery et al. , 2011; Silva et al. , 2017).

India has a national ambient surface monitoring network that started in 1987 and has become more extensive over time with a substantial increase in the number and spatial extent of continuous and manual monitoring stations between 2015 and 2019. At present, the Central Pollution Control Board (CPCB), along with the State Pollution Control Boards (SPCBs), run the most extensive monitoring network in the country under the National Air Quality Monitoring Program (NAMP). As of 2019, NAMP cooperatively operated (with CPCB and SPCBs) over 750 manual monitoring stations (compared with 20 in 1987 when monitoring first began and 450 in 2015 when our analysis starts) which publicly archive annual average concentrations of PM 10 , PM 2.5 , SO 2 and NO 2 ( https://cpcb.nic.in/namp-data/ ). As of 2019, over 220 Continuous Ambient Air Quality Monitoring (CAAQM) stations operated (compared with less than 50 stations in 2015 when our analysis starts). CPCB archives publicly available, real time data, every 15 minutes, from over 220 stations across India of an extensive list of criteria and non-criteria air pollutants and meteorological variables ( https://app.cpcbccr.com/ccr/ ). Stations vary in the air pollutant species and meteorological data they collect. The manual monitors provide better spatial coverage than the continuous monitors but provide data on fewer air pollutants at much lower temporal resolution (annual average values versus every 15 minutes). However, both sets of monitoring stations sample exclusively urban areas despite the fact that rural areas have significant emissions from households and agricultural waste burning (Balakrishnan et al. , 2014; Venkatraman et al. , 2018). Pant et al. (2019) and the Supplementary Information (SI) (Section 1) describe other Indian monitoring networks which are less extensive and are not publicly available. India has fewer monitoring stations than most south and east Asian countries, with ~1 monitor/6.8 million persons (Apte and Pant 2019; Brauer et al. , 2019; Martin et al. , 2019). Despite recent increases in urban monitoring stations across India, vast regions do not have monitors and except for satellite data for a few species, little information is available on surface concentrations of air pollutants in non-urban locations in India.

Recently, extreme levels of fine particulate air pollution in India, combined with a growing appreciation of the adverse impacts of elevated air pollution on health, led the Indian government to launch the National Clean Air Program (NCAP) in 2019 (Ministry of Environment, Forests and Climate Change NCAP, 2019). NCAP targets a reduction of 20–30% in PM 10 and PM 2.5 concentrations by 2024 relative to 2017 levels. One focus of NCAP is augmentation of the national monitoring network for which substantial financial support was announced in the 2020 Union Budget.

Despite a growing monitoring network and the need for analysis, prior to our work, no study holistically analyzed existing government surface air pollutant monitoring data across India. Most research studies analyzing ground monitoring data have focused on Delhi and the surrounding National Capital Region (NCR) (Guttikunda and Gurjar, 2012; Sahu and Kota, 2017; Sharma et al. , 2018; Chowdhury et al. , 2019; Guttikunda et al. , 2019; Wang and Chen, 2019; Hama et al. , 2020), and other major cities (Gurjar et al. , 2016; Sreekanth et al. , 2018, Yang et al. , 2018; Chen et al. , 2020). In addition, some studies also used ground observations to bias correct satellite measurements for India (Pande et al. , 2018; Chowdhury et al. , 2019; Navinya et al. , 2020). However, a need remains for a comprehensive analysis of all surface data collected by manual NAMP and continuous CAAQM monitoring networks between 2015–2019 over which period monitoring increased substantially.

Here we provide the first national analysis of all available surface measurements of key criteria pollutants (PM 10 , PM 2.5 , SO 2 , NO 2 and O 3 ) across India between 2015–2019. We use publicly available data from the NAMP manual and CAAQM real-time stations which have different spatial distributions and temporal resolutions. Collating spatio-temporal distributions of pollutant concentrations on inter-annual, annual, seasonal and monthly timescales, we present an overview of the variability in air pollution levels across the country and separately analyze pollution levels in northern (north of 23°N) and southern India. We conduct case studies of five cities in India in which U.S. State Department PM 2.5 monitors (Air-Now network) are present and, using additional data collected by CAAQM monitors, compare pollution status between these cities. We also compare analyzed annual average PM 2.5 from the CAAQM network with the satellite derived surface PM 2.5 (Hammer et al. , 2020) and find good agreement between the two datasets. Our analysis will provide a valuable baseline to evaluate the future success of the NCAP in meeting its air pollution mitigation targets.

2 METHODOLOGY

2.1 criteria pollutant data.

We analyze all open-source data available from the manual (NAMP) and continuous (CAAQM) networks, as well as from the US Embassy and consulates Air-Now network from 2015–2019 for five criteria pollutants—PM 10 , PM 2.5 , SO 2 , NO 2 and O 3 .

Datasets from 2015-2018 were acquired for NAMP and were acquired from 2015–2019 for CPCB-CAAQM and Air-Now networks directly from the following sources:

NAMP manual monitoring network ( https://cpcb.nic.in/namp-data/ ): Annual average and annual maximum and minimum concentrations were obtained from a total of 730 manual stations. Higher resolution temporal measurements are not publicly reported by NAMP. We analyze data from 2015–2018 as datasets for 2019 were unavailable when our analysis was completed in December 2020.
CAAQM continuous monitoring network from the Central Control Room for Air Quality Management website ( https://app.cpcbccr.com/ccr/ ): One-hour averages were calculated from reported 15 minute average concentrations. Neither the continuous nor manual monitoring stations include geolocations. To obtain the latitude/longitude coordinates of each station, we used the monitoring station name and geolocated them using Google maps.
S. State Department Air-Now network ( https://www.airnow.gov/ ): One-hour average PM 2.5 concentrations were obtained for monitors located in Delhi, Mumbai, Hyderabad, Kolkata and Chennai.

2.2 Data Quality Control

We directly utilize the data available from the NAMP and Air-Now networks, but process the data we use from the CAAQM network to ensure representative monthly, seasonal, and annual average air pollutant concentrations using the following method:

Missing data is removed. Values in excess of the reported range (see Table S1) are assumed to be errors and are removed. Values of 999.99 for PM 10 and PM 5 are retained as they may represent concentrations above the upper detection limit of the instrument. The U.S. Air-Now network data in New Delhi report 1-hour average PM 2.5 concentrations between 1300 and 1486 µg m – 3 during Diwali for each year. As CAAQM does not report values in excess of 999.99 µg m – 3 for PM 2.5 our annual means based on CAAQM will likely be biased low in some locations. In sequences of 24 or more consecutive identical hourly values, only the first value out of the sequence is retained. Data were processed following the QA/QC procedure described below. The percentage of data removed due to this processing is provided in Tables S2(a) and S2(b).
Diurnal mean values are calculated for criteria pollutants PM 10 , PM 5 , SO 2 , NO 2 and O 3 for each 12-hour day-night interval (between 6 am–6 pm and 6 pm–6 am (next day)), using a minimum of one hourly observation for each 12-hour period. Daily means are calculated only for days that have a daytime or nighttime mean value. For O 3 , daily mean (MDA8) values are calculated as the maximum of 8-hour moving averages over a 24-hour period using at least 6 hourly observations. For all pollutants, monthly mean values are calculated for months that have at least 8 daily mean values (at least 25% of observations). To obtain annual average concentrations, we calculate quarterly means and require at least one monthly mean value as input to each quarterly mean concentration. At least two quarterly mean values are used for calculating annual average concentrations. This procedure is followed to ensure representativeness of data in diurnal, daily, monthly, seasonal, annual and interannual timeseries. Fig. 1 shows a flow chart describing the methodology for generating each step of the time-series.

Fig. 1. Methodology used to create a representative data series for each pollutant which provides daily, monthly, seasonal and annual average concentrations.

3 RESULTS

3.1 strengths and weaknesses of available air quality datasets.

Until the start of 2018 the Indian monitoring network had limited extent. Very few stations have operated continuously from 2015 to the present. The number of stations in the continuous monitoring network has increased dramatically since 2017 ( Fig. 2 ) making it far more feasible now to evaluate air quality across India than in the past. However, spatial coverage is still limited with unequal distribution of monitors. All monitors are in cities, with a concentration in the largest cities, and none are in rural areas. Fig. 3 shows the percentage of valid hourly observations, compared with total hours annually, from each CAAQM station between 2015 and 2019. Although the current data is sufficient to provide an overview of air quality across much of India, it is currently challenging to use air quality datasets to conduct long term trend analysis due to their limited spatial and temporal coverage.

Fig. 2. Number of CAAQM stations providing valid hourly concentrations across India, between 2015–2019, for PM10, PM2.5, SO2, NO2 and O3, respectively.

3.2 Spatial Distribution of Air Pollutants from 2015–2019

Figs. 4 and 5 show annual average concentrations of five criteria pollutants (PM 10 , PM 2.5 , SO 2 , NO 2 and O 3 ) at continuous and manual monitoring stations across India, from 2015 to 2019. The general distribution pattern of air pollution, showing higher pollution levels in northern than southern India, is captured in both the manual and continuous monitoring station data.

Fig. 4. Spatial distribution of annual average (2015–2019) concentrations (µg m–3) of PM10, PM2.5, SO2, NO2 and maximum daily average 8-hour (MDA8) O3 from the CPCB CAAQM continuous monitoring stations that meet our criteria for data inclusion (see methods for details). Each dot represents a single station. The number of stations for each species in each year is indicated in parentheses.

The number of continuous and manual monitoring stations have both increased substantially between 2015 and 2019 with 15 (147) CAAQM stations meeting our criteria for PM 10 , 33 (181) for PM 2.5 , 31 (163) for SO 2 , 34 (175) for NO 2 and 32 (168) for O 3 and in 2015 (2019) (see Figs. 4 and 5 for details of other years and manual stations). Of the total, nearly 60% of the CAAQM continuous monitoring stations are in northern India with 20% of the total stations in Delhi in 2019. Despite being a high pollution zone with nearly 15% of the Indian population ( http://up.gov.in/upstateglance.aspx ), the Indo Gangetic Plain has only 13% (9%) of total continuous (manual) monitoring stations. NAMP manual monitoring stations are more widely distributed than continuous monitors across India, with more monitors in the south and thus provide more representative spatial distributions of pollutants. However, they only provide annual average pollutant concentrations and thus cannot be used to analyze seasonal variations.

Elevated concentrations of PM 10 and PM 2.5 were recorded by both CAAQM and NAMP manual monitors across northern Indian states in all years, with particularly high concentrations across the Indo-Gangetic Plain (IGP). Ground observations of SO 2 are generally low across the country with high concentrations found at a few urban and industrial locations. This has been corroborated by previous studies (Guttikunda and Calori, 2013). The role of alkaline dust in scavenging SO 2 in India likely reduces ambient concentrations (Kulshrestha et al. , 2003). In contrast, annual average NO 2 and MDA8 O 3 concentrations are highly variable depending on location with higher O 3 concentrations often seen in the IGP region.

3.3 Annual Variation in Pollutant Concentrations in Northern and Southern India

The spatial distribution of pollutants is affected by meteorology, geography, topography, population density, location specific emission sources including industries, vehicular density, resuspended dust from poor land use management etc. In northern India (north of 23.5°N), higher population density and higher associated activities in industry, transport, power generation, seasonal crop residue burning, and more frequent dust storms contribute to higher particulate loads than in southern India (Sharma and Dixit, 2016; Cusworth et al. , 2018). We observed significant differences between northern and southern India in the spatio-temporal patterns of PM 10 , PM 2.5 , SO 2 , NO 2 and MDA8 O 3 .

Fig. 6 shows annual average concentrations (µg m – 3 ) of PM 10 , PM 2.5 , SO 2 , NO 2 and MDA8 O 3 respectively, for northern and southern India (divided at 23.5°N) from CAAQM stations. The number of stations used to calculate annual average values is shown in Fig. 4 for each species. Annual average concentrations of PM 10 , PM 2.5 , and NO 2 are higher in northern India, whereas SO 2 and MDA8O 3 are similar in the north and the south. Annual average concentrations from CAAQM continuous and NAMP manual monitoring stations, combined (S1 a), and only manual monitoring Stations (S1 b) are plotted separately in Fig. S1. We found inter-annual variability but no significant annual trend in the timeseries of these pollutants. Annual average concentrations over the five year period in northern (and southern) India were: 197 ± 84 µg m – 3 (93 ± 30 µg m – 3 ) for PM 10 , 109 ± 29 µg m – 3 (47 ± 16 µg m – 3 ) for PM 2.5 , 12 ± 7 µg m – 3 (12 ± 10 µg m – 3 ) SO 2 , 35 ± 21 µg m – 3 (27 ± 16 µg m – 3 ) for NO 2 and 73 ± 29 µg m – 3 (66 ± 31 µg m – 3 ) for MDA8 O 3 . In the five-year period, annual NAAQS were met at approximately 3% of all CAAQM stations measuring PM 10 , 13% of PM 2.5 , 70% of NO 2 and 98% of SO 2 (Table S3). MDA8 O 3 standard of 100 µg m – 3 (to be met 98% of the time within a year) was met at 77% of all CAAQM stations between 2015–2019, inclusive. Particulate matter dominates the pollution mix with national average annual mean concentrations exceeding the NAAQ standard for all analyzed years and in northern India more than double the allowed concentration. Fig. 7 shows annual average concentrations of these pollutants from CAAQM stations that meet our analysis criteria and are available each year from 2015 through 2019. The change in annual concentrations relative to the annual average concentrations in 2015–2017 at the stations operational throughout this period is shown in Fig. S2 in order to provide a comparison useful for evaluating the success of the NCAP.

Fig. 6. Annual average concentrations (µg m–3) of PM10, PM2.5, SO2, NO2 and MDA8 O3 from all CAAQM continuous stations from 2015 through 2019, for northern and southern India (divided at 23.5°N and shown in left and right panels). Box edges indicate the interquartile range, whiskers indicate the maximum and minimum values, dashed lines inside the box are the medians and colored triangles indicate annual mean concentrations. CPCB and WHO ambient air quality standards are shown in magenta and blue dotted lines, respectively. Annual standards are provided for PM10, PM2.5, NO2 and SO2. (WHO does not provide an annual SO2 ambient air quality standard. It provides a 24-hour average standard of 40 µg m–3). For O3, maximum daily average 8-hour (MDA8) O3 standard is mentioned. (CPCB air quality standards apply to industrial, residential, rural and other areas. Ecologically sensitive areas have different standards and are not included).

3.5 Seasonal and Monthly Patterns of Air Pollutants

Seasonal concentrations of air pollutants in India are heavily influenced by meteorology and location. Influence of meteorology on spatio-temporal distributions of pollutants across India is described in Section S3. Fig. S3 shows the mean seasonal distribution of boundary layer height, surface pressure, precipitation, and omega/vertical and horizontal wind velocity. We calculate seasonal and monthly concentrations of PM 10 , PM 2.5 , SO 2 , NO 2 and MDA8 O 3 between 2015–2019 for northern and southern India in each season ( Fig. 8 ) and month ( Fig. 9 ) and show seasonal spatial distributions of these pollutants across India (Fig. S4). We analyze seasonal composites computed as averages for the spring or pre-monsoon period, March–April–May (MAM), the monsoon period, June–July–August (JJA), the autumn or post monsoon period, September–October–November (SON) and winter, December–January–February (DJF). In all seasons, substantially higher concentrations are observed for PM 10 and PM 2.5 , in northern India with concentrations of NO 2 , SO 2 and MDA8 O 3 only slightly more elevated in northern than southern India. The DJF average concentrations are highest for PM 10 , PM 2.5 and NO 2 in northern (southern) India: 270 ± 51 (137 ± 11) µg m –3 , 170 ± 26 (69 ± 2) µg m –3 , 47 ± 2 (35 ± 7) µg m –3 , respectively. Seasonal average concentrations of SO 2 peak in MAM in northern India (15 ± 3 µg m –3 ) and in DJF in southern India (16 ± 4 µg m –3 ), with highest concentrations in winter across the country. For DA8 O 3 , highest seasonal concentrations occur in MAM (DJF) in the north 71.8 ± 28 µg m –3 and south (84 ± 8 µg m –3 ).

Fig. 8. Seasonal average concentrations for northern (solid lines) and southern India (dashed lines) (divided at 23.5°N latitude) from 2015–2019, inclusive, of PM10, PM2.5, SO2, NO2 and MDA8 O3 (µg m–3) from all CAAQM stations meeting analysis criteria. See Fig. 4 for station locations and annual average concentrations.

Monthly variations in pollution are also a function of regional circulation patterns. The summer monsoon facilitates dilution of pollution via strong south-westerly winds from the Arabian Sea and wet scavenging of anthropogenic pollution (Zhu et al. , 2012). Wet deposition removes PM 10 , PM 2.5 and water soluble SO 2 (Chin, 2012) leading to substantially lower ambient concentrations of these pollutants in JJA across India. Minimum concentrations of all pollutants occur in August.

Outside the monsoon, weak regional circulation and large scale high pressure systems result in accumulation of pollutants near the surface which is most pronounced in winter. Highest monthly concentrations are seen in November–January, inclusive, for PM 10 , PM 2.5 , SO 2 and NO 2 . For, MDA8O 3 , highest monthly concentrations are recorded in May (January) for northern (southern) India. Precursor emissions, surface temperature and solar insolation modulate a complex chemistry that drives the ozone cycle (Lu et al. , 2018).

3.6 Case studies of Delhi, Kolkata, Mumbai, Hyderabad and Chennai

Delhi, Kolkata, Mumbai, Hyderabad and Chennai are the five cities in India in which the U.S. State Department Air-Now network real time monitoring stations record PM 2.5 concentrations at the US embassy and consulates. In these five cities, we compare daily and monthly mean PM 2.5 measurements from the Air-Now and CAAQM networks. Fig. 10 shows scatterplots between daily mean PM 2.5 from the Air-Now monitor located in each of the five cities with all CPCB CAAQM monitors in those cities for 2015–2019, inclusive. We find a good correlation between the daily average PM 2.5 concentrations from the two networks at all the cities (r > 0.8), except Chennai (r~0.47) where CPCB concentrations are biased higher than the Air-Now concentrations. On highly polluted days in Delhi, the Air-Now monitors report higher PM 2.5 concentrations than the CPCB monitors in part because Air-Now monitors are able to report hourly concentrations above 1000 µg m –3 while the CPCB monitors cannot.

Fig. 10. Scatter plots of daily mean PM2.5 concentrations comparing Air-Now observations from the five cities in which they exist with all CPCB CAAQM monitors in those cities, between 2015–2019. For each plot the regression line (solid), regression equation and r value for each correlation are shown for each city. The dashed grey line indicates 1:1 correspondence. The inset plots are scaled to the data range.

We examine how concentrations of PM 10 , PM 2.5 , SO 2 , NO 2 and O 3 vary between cities in which Air-Now monitors exist from 2015–2019 (see Fig. 11). Fig. 11 compares the monthly average concentrations of PM 2.5 between the two networks, examines the variation in concentrations over time for other species measured only by CPCB, and compares observed concentrations with the annual NAAQS for residential areas. Annual average concentrations from the stations combined in each city that meet our criteria is shown in Fig. S5 and a timeseries for each pollutant at each station is shown in Fig. S6. From CAAQM and Air-Now networks, we find Delhi has the highest daily, monthly mean and annual average concentrations of PM 10 and PM 2.5 , followed by Kolkata and Mumbai (Figs. 10, 11; Fig. S5).

Fig. 11. Timeseries of monthly mean concentrations in Delhi, Kolkata, Mumbai, Hyderabad and Chennai (north to south order) of PM2.5 (CPCB CAAQM and Air-Now network) and PM10, NO2, SO2 and MDA8 O3 from all CAAQM stations in the five cities from 2015 to 2019 meeting our analysis criteria. The dots represent monthly means and the shaded region, in the same color as the dots, indicates values within one standard deviation of the mean for each city. Values following the station names indicate the number of monitoring stations included in the analysis of each city. Annual average residential area NAAQS for each pollutant are shown with a dashed black line (PM10 = 60 µg m–3, PM2.5 = 40 µg m–3; SO2 = 50 µg m–3; NO2 = 40 µg m–3; MDA8 O3 = 100 µg m–3 (not to be exceeded more than 2% of the year)).

For Delhi, between 2015 and 2019, annual average concentrations of PM 2.5 from the CAAQM station closest to the U.S. embassy (RK Puram, Delhi) greatly exceeded the residential NAAQS for PM 2.5 of 40 µg m –3 and ranged from 101 to 119 µg m –3 with the Air-Now station ranging from 95 to 124 µg m –3 . Chennai has the lowest monthly and annual average concentrations of PM 2.5 . The US state department annual average PM 2.5 values overall are consistent with the CAAQM stations and show a similar trend across cities. All five cities failed to meet the annual average CPCB PM 10 standard of 60 µg m –3 in all years.

Monthly and annual average SO 2 concentrations are far below the annual standard of 50 µg m –3 at all locations throughout the year in these five cities with Delhi reporting the highest annual average concentrations among the five cities followed by Mumbai. Starting in 2018 both Delhi and Mumbai had SO 2 concentrations lower than prior years.

Monthly average NO 2 concentrations are highest in Delhi in all years and starting in 2017, decrease from a peak over 100 µg m –3 in 2017 to a peak of 52 µg m –3 in 2019. Kolkata and Hyderabad also have relatively high concentrations of NO 2 with annual average concentrations exceeding the residential NAAQS of 40 µg m –3 starting in 2018.

Monthly MDA8 O 3 concentrations across all five cities are similar, particularly after 2018 and are generally falling below the residential 8-hour average NAAQS of 100 µg m 3 . Similar monthly tropospheric ozone concentrations in these cities, despite different levels of particulate matter, NO 2 and meteorology, make it a topic for further investigation.

4 DISCUSSION

4.1 growing dataset and existing gaps.

Prior to 2015 surface air quality monitoring data was available from only a few stations in India. Over the period we analyzed, 2015–2019, the number of monitoring stations across India increased dramatically. Our compilation and rigorous quality control of these data provide, for the first time, a comprehensive dataset of criteria pollutants that can be used to evaluate air pollutant concentrations simulated by atmospheric chemical transport models, satellite retrievals and reanalysis. Our dataset also provides a baseline for the NCAP. Previous studies have used ground observations from selected locations without transparently addressing existing data gaps and are not clear in their evaluation and quality assurance of surface observations. Here, we have carefully evaluated the archived data for completeness and accuracy, discarding values in excess of instrumental range, and requiring representative temporal coverage for each averaging period at each monitor. For example, for inclusion in our analysis a monitor measuring a species we analyze must report daily averages at least one hour per 12-hour daytime or night-time period, eight days for each monthly average, and one month per quarter and atleast two quarters for each annual average (see Tables S2(a), S2(b) and S3). However, spatial coverage remains spotty with monitoring stations predominantly located in large cities; smaller cities and rural locations lack coverage. Further expansion of the monitoring networks to facilitate an improved understanding of spatial distributions of pollutants across urban/rural India and to evaluate future trends in pollutant concentrations is needed. Very few stations provide valid observations continuously from 2015 onwards limiting our ability to analyze past trends in air quality. However, trend analyses starting in 2018 will be valuable and possible in the future.

4.2 Differences in Air Quality Observations

We compare monthly, seasonal and annual mean concentrations of air pollutants we analyze with other studies that have analyzed surface measurements of the same pollutants, cities and time periods across India (Table S5). We find that the range of concentrations of criteria pollutants reported in our analysis of CPCB data are similar to the values presented in research studies using ground observations during the same period (Kota et al. , 2018; Sreekanth et al. , 2018; Guttikunda et al. , 2019; Mahesh et al. , 2019; Ravinder et al. , 2019; Jain et al. , 2020; Tyagi et al. , 2020; Jat et al. , 2021). However, as shown in Table S5, in case studies covering extreme events and studies in bigger cities and more polluted regions, like Delhi and the IGP, differences exist between the CPCB concentrations we calculate and those reported in the literature from surface monitoring stations, models and satellite data (Kota et al. , 2018; Tyagi et al. , 2019; Jat et al. , 2021).

In Fig. 12 , we compare the spatial patterns of annual average surface PM 2.5 concentrations derived from satellite data with measurements from the CPCB continuous network. The surface satellite concentrations were obtained by combining data from Aerosol Optical Depth (AOD) from MODIS (Moderate Resolution Imaging Spectroradiometer), MISR (Multi-angle Imaging Spectroradiometer), MAIAC (Multi Angle Implementation of Satellite Correction) and SeaWiFS (Sea Viewing Wide Field of View Sensor) satellite products and using the GEOS-Chem model to obtain gridded surface PM 2.5 concentrations at 0.05° × 0.05° (Hammer et al. , 2020). The product we use is V4.GL.03 available at https://sites.wustl.edu/acag/datasets/surface-pm2-5/#V4.GL.03 . Reasonable agreement is seen between the annual mean surface concentrations of PM 2.5 derived from the satellite data and from the CPCB CAAQM observations from 2015-2019. Agreement is particularly good over the IGP and in central and southern India. However, along the western desert region (near Thar desert in Rajasthan), satellite concentrations of surface PM 2.5 (~40–50 µg m –3 ) were substantially lower than concentrations obtained from the CPCB CAAQM monitors (~80–100 µg m –3 ) for 2015–2017. In 2018 and 2019 the correspondence between the two datasets improved with most annual mean PM 2.5 concentrations in the western desert region generally between ~40 and 60 µg m –3 .

Fig. 12. Satellite derived annual surface PM2.5 concentration overlaid with CAAQM network surface measurements (circles), from 2015–2019.

5 CONCLUSIONS

This study provides the first comprehensive analysis of all existing government monitoring data available for PM 10 , PM 2.5 , SO 2 , NO 2 and MDA8 O 3 using the continuous (CAAQM) and manual (NAMP) monitoring networks in India as well as the data from the US State Department Air-Now network, between 2015 and 2019 (2018 for NAMP). Our analysis shows that the Indian data record, in terms of number of monitoring stations, observations and quality of data, has improved significantly over this period. Despite the effort to augment surface monitoring infrastructure, gaps remain in spatial and temporal coverage and additional monitoring stations in small cities and rural areas are needed. Monitoring stations located in bigger cities (e.g., five Air-Now cities) have better data quality, from more widely distributed stations within the city, than is available for smaller cities. Pollution hotspots are occasionally found in smaller cities where monitoring stations are sparse. No stations have yet been placed in rural areas and are needed there in order to better characterize air quality and pollution sources across India (e.g., the effect of agricultural waste burning on air quality).

We find that fine particulate pollution dominates the pollution mix across India with virtually all sites in northern India (north of 23.5°N) exceeding the annual average PM 10 and PM 2.5 national residential ambient air quality standards (NAAQS) by 150% and 100% respectively, and in southern India (south of 23.5°N) exceeding the PM 10 standard by 50% and PM 2.5 standard by 40%. Comparison of PM 2.5 surface observations from the CPCB continuous monitoring network with surface satellite concentrations finds good agreement across India, particularly for 2017 and 2018. Prior to 2017 CAAQM concentrations were substantially higher than indicated by the satellite data over the western desert region. Annual average SO 2 , NO 2 and MDA8 O 3 generally meet the residential NAAQS across India. We find that northern India has (~10%–130%) higher average concentrations of all pollutants than southern India, except for SO 2 where the concentrations are similar. Although inter-annual variability exists, no significant trend of these pollutants was observed over the five-year period except for a small decrease over time in PM 10 and PM 2.5 in winter, which is more pronounced in the stations in northern and central India.

Our analysis of surface measurements is valuable for evaluating air pollutant concentrations simulated in atmospheric chemistry models. We found good agreement between the annual average CAAQM PM 2.5 we analyzed and satellite derived surface PM 2.5 from Hammer et al. (2020). Our data set can also be used to evaluate satellite retrievals of NO 2 and O 3 as well as seasonal variability in PM 2.5 concentrations. Finally, India is targeting a reduction of 20–30% in particulate pollution under NCAP by 2024 relative to 2017. Our analysis from 2015–2019 at different spatial and temporal scales of surface pollution provides a baseline to evaluate the future success of the programme as well as aids in the assessment of existing and future air pollution mitigation policies.

ADDITIONAL INFORMATION

data access.

The raw data from the continuous CPCB monitors used in our analyses along with the code for data quality control and the calculation of various temporal averages is available at https://doi.org/10.34770/60j3-yp02

ACKNOWLEDGEMENTS

We thank Mi Zhou for early assistance in data processing and two anonymous reviewers for helpful suggestions to improve our manuscript. Funding for D.S. was provided by a Science, Technology and Environmental Policy fellowship at the Center for Policy Research on Energy and Environment at Princeton University.

Apte, J.S., Pant, P. (2019). Toward cleaner air for a billion Indians. PNAS 116, 10614–10616. https://doi.org/10.1073/pnas.1905458116
Avnery, S., Mauzerall, D.L., Fiore, A.M. (2013). Increasing global agricultural production by reducing ozone damages via methane emission controls and ozone‐resistant cultivar selection. Global Change Biol. 19, 1285–1299. https://doi.org/10.1111/gcb.12118
Avnery, S., Mauzerall, D.L., Liu, J., Horowitz, L.W. (2011). Global crop yield reductions due to surface ozone exposure: 2. Year 2030 potential crop production losses and economic damage under two scenarios of O 3 pollution. Atmos. Environ. 45, 2297–2309. https://doi.org/10.1016/j.atmosenv.2011.01.002
Balakrishnan, K., Cohen, A., Smith, K.R. (2014). Addressing the burden of disease attributable to air pollution in India: The need to integrate across household and ambient air pollution exposures. Environ. Health Perspect. 122. https://doi.org/10.1289/ehp.1307822
Balakrishnan, K., Dey, S., Gupta, T., Dhaliwal, R.S., Brauer, M., Cohen, A.J., Stanaway, J.D., Beig, G., Joshi, T.K., Aggarwal, A.N., Sabde, Y., Sadhu, H., Frostad, J., Causey, K., Godwin, W., Shukla, D.K., Kumar, G.A., Varghese, C.M., Muraleedharan, P., Agrawal, A., et al. (2019). The impact of air pollution on deaths, disease burden, and life expectancy across the states of India: The Global Burden of Disease Study 2017. Lancet Planet. Heath 3, e26–e39. https://doi.org/10.1016/S2542-5196(18)30261-4
Brauer, M., Guttikunda, S.K., Nishad, K.A., Dey, S., Tripathi, S.N., Weagle, C., Martin, R.V. (2019). Examination of monitoring approaches for ambient air pollution: A case study for India. Atmos. Environ. 216, 116940. https://doi.org/10.1016/j.atmosenv.2019.116940
Chen, Y., Wild, O., Conibear, L., Ran, L., He, J., Wang, L., Wang, Y. (2020). Local characteristics of and exposure to fine particulate matter (PM 2.5 ) in four Indian megacities. Atmos. Environ. 5, 100052. https://doi.org/10.1016/j.aeaoa.2019.100052
Chin, M. (2012). Dirtier air from a weaker monsoon. Nat. Geosci. 5, 449–450. https://doi.org/10.1038/ngeo1513
Chowdhury, S., Dey, S. (2016). Cause-specific premature death from ambient PM 2.5 exposure in India: Estimate adjusted for baseline mortality. Environ. Int. 91, 283–290. https://doi.org/10.1016/j.envint.2016.03.004
Chowdhury, S., Dey, S., Di Girolamo, L., Smith, K.R., Pillarisetti, A., Lyapustin, A. (2019). Tracking ambient PM 2.5 build-up in Delhi national capital region during the dry season over 15 years using a high-resolution (1 km) satellite aerosol dataset. Atmos. Environ. 204, 142–150. https://doi.org/10.1016/j.atmosenv.2019.02.029
Chowdhury, S., Pozzer, A., Dey, S., Klingmueller, K., Lelieveld, J. (2020). Changing risk factors that contribute to premature mortality from ambient air pollution between 2000 and 2015. Environ. Res. Lett. 15, 074010. https://doi.org/10.1088/1748 9326/ab8334
Cusworth, D.H., Mickley, L.J., Sulprizio, M.P., Liu, T., Marlier, M.E., DeFries, R.S., Guttikunda, S.K., Gupta, P. (2018). Quantifying the influence of agricultural fires in northwest India on urban air pollution in Delhi, India. Environ. Res. Lett. 13, 044018. https://doi.org/10.1088/1748-9326/aab303
Gao, M., Gao, J., Zhu, B., Kumar, R., Lu, X., Song, S., Zhang, Y., Jia, B., Wang, P., Beig, G., Hu, J., Ying, Q., Zhang, H., Sherman, P., McElroy, M.B. (2020). Ozone pollution over China and India: Seasonality and sources. Atmos. Chem. Phys. 20, 4399–4414. https://doi.org/10.5194/acp-20-4399-2020
GDB Maps Working Group (2018). Burden of disease attributable to major air pollution sources in India.
Ghude, S.D., Jena, C., Chate, D.M., Beig, G., Pfister, G.G., Kumar, R., Ramanathan, V. (2014). Reductions in India's crop yield due to ozone. Geophys. Res. Lett. 41, 5685–5691. https://doi.org/10.1002/2014GL060930
Gurjar, B.R., Ravindra, K., Nagpure, A.S. (2016). Air pollution trends over Indian megacities and their local-to-global implications. Atmos. Environ. 142, 475–495. https://doi.org/10.1016/j.atmosenv.2016.06.030
Guttikunda, S.K., Gurjar, B.R. (2012). Role of meteorology in seasonality of air pollution in megacity Delhi, India. Environ. Monit. Assess. 184, 3199–3211. https://doi.org/10.1007/s10661-011-2182-8
Guttikunda, S.K., Calori, G. (2013). A GIS based emissions inventory at 1 km× 1 km spatial resolution for air pollution analysis in Delhi, India. Atmos. Environ. 67, 101–111. https://doi.org/10.1016/j.atmosenv.2012.10.040
Guttikunda, S.K., Nishadh, K.A., Jawahar, P. (2019). Air pollution knowledge assessments (APnA) for 20 Indian cities. Urban Clim. 27, 124–141. https://doi.org/10.1016/j.uclim.2018.11.005
Hama, S.M.L., Kumar, P., Harrison, R.M., Bloss, W.J., Khare, M., Mishra, S., Namdeo, A., Sokhi, R., Goodman, P., Sharma, C. (2020). Four-year assessment of ambient particulate matter and trace gases in the Delhi-NCR region of India. Sustainable Cities Soc. 54, 102003. https://doi.org/10.1016/j.scs.2019.102003
Hammer, M.S., van Donkelaar, A., Li, C., Lyapustin, A., Sayer, A.M., Hsu, N.C., Levy, R.C., Garay, M.J., Kalashnikova, O.V., Kahn, R.A., Brauer, M., Apte, J.S., Henze, D.K., Zhang, L., Zhang, Q., Ford, B., Pierce, J.R., Martin, R.V. (2020). Global estimates and long-term trends of fine particulate matter concentrations (1998–2018). Environ. Sci. Technol. 54, 7879–7890. https://doi.org/10.1021/acs.est.0c01764
Jain, S., Sharma, S.K., Vijayan, N., Mandal, T.K. (2020). Seasonal characteristics of aerosols (PM 2.5 and PM 10 ) and their source apportionment using PMF: A four year study over Delhi, India. Environ. Pollut. 262, 114337. https://doi.org/10.1016/j.envpol.2020.114337
Jat, R., Gurjar, B.R., Lowe, D. (2021). Regional pollution loading in winter months over India using high resolution WRF-Chem simulation. Atmos. Res. 249, 105326. https://doi.org/10.1016/j.atmosres.2020.105326
Kota, S.H., Guo, H., Myllyvirta, L., Hu, J., Sahu, S.K., Garaga, R., Ying, Q., Gao, A., Dahiya, S., Wang, Y., Zhang, H. (2018). Year-long simulation of gaseous and particulate air pollutants in India. Atmos. Environ. 180, 244–255. https://doi.org/10.1016/j.atmosenv.2018.03.003
Kulshrestha, M.J., Kulshrestha, U.C., Parashar, D.C., Vairamani, M. (2003). Estimation of SO 4 contribution by dry deposition of SO 2 onto the dust particles in India. Atmos. Environ. 37, 3057–3063. https://doi.org/10.1016/S1352-2310(03)00290-5
Lu, X., Zhang, L., Liu, X., Gao, M., Zhao, Y., Shao, J. (2018). Lower tropospheric ozone over India and its linkage to the South Asian monsoon. Atmos. Chem. Phys. 18, 3101–3118. https://doi.org/10.5194/acp-18-3101-2018
Mahesh, B., Rama, B.V., Spandana, B., Sarma, M.S.S.R.K.N., Niranjan, K., Sreekanth, V. (2019). Evaluation of MERRAero PM 2.5 over Indian cities. Adv. Space Res. 64, 328–334. https://doi.org/10.1016/j.asr.2019.04.026
Martin, R.V., Brauer, M., van Donkelaar, A., Shaddick, G., Narain, U., Dey, S. (2019). No one knows which city has the highest concentration of fine particulate matter. Atmos. Environ. 3, 100040. https://doi.org/10.1016/j.aeaoa.2019.100040
Ministry of Environment, Forests and Climate Change (MoEF) (2019). National Clean Air Program (NCAP). Government of India. http://moef.gov.in/wp-content/uploads/2019/05/NCAP_Report.pdf
National Ambient Air Quality Monitoring Program (NAMP) (2020). Data Year wise. https://cpcb.nic.in/namp-data/ (accessed 15 December 2020).
Navinya, C.D., Vinoj, V., Pandey, S.K. (2020). Evaluation of PM 2.5 surface concentrations simulated by NASA’s MERRA Version 2 aerosol reanalysis over India and its relation to the air quality index. Aerosol Air Qual. Res. 20, 1329–1339. https://doi.org/10.4209/aaqr.2019.12.0615
Pande, P., Dey, S., Chowdhury, S., Choudhary, P., Ghosh, S., Srivastava, P., Sengupta, B. (2018). Seasonal transition in PM 10 exposure and associated all-cause mortality risks in India. Environ. Sci. Technol. 52, 8756–8763. https://doi.org/10.1021/acs.est.8b00318
Pandey, A., Brauer, M., Cropper, M.L., Balakrishnan, K., Mathur, P., Dey, S., Turkgulu, B., Kumar, G.A., Khare, M., Beig, G., Gupta, T., Krishnankutty, R.P., Causey, K., Cohen, A.J., Bhargava, S., Aggarwal, A.N., Agrawal, A., Awasthi, S., Bennitt, F., Bhagwat, S., et al. (2021). Health and economic impact of air pollution in the states of India: The Global Burden of Disease Study 2019. Lancet Planet. Heath 5, e25–e38. https://doi.org/10.1016/S2542-5196(20)30298-9
Pant, P., Lal, R.M., Guttikunda, S.K., Russell, A.G., Nagpure, A.S., Ramaswami, A., Peltier, R.E. (2019). Monitoring particulate matter in India: recent trends and future outlook. Air Qual. Atmos. Health 12, 45–58. https://doi.org/10.1007/s11869-018-0629-6
Ravindra, K., Singh, T., Mor, Sahil, Singh, V., Mandal, T.K., Bhatti, M.S., Gahlawat, S.K., Dhankhar, R., Mor, Suman, Beig, G. (2019). Real-time monitoring of air pollutants in seven cities of North India during crop residue burning and their relationship with meteorology and transboundary movement of air. Sci. Total Environ. 690, 717–729. https://doi.org/10.1016/j.scitotenv.2019.06.216
Sahu, S.K., Kota, S.H. (2017). Significance of PM 2.5 air quality at the Indian capital. Aerosol Air Qual. Res. 17, 588–597. https://doi.org/10.4209/aaqr.2016.06.0262
Sharma, M., Dikshit, O. (2016). Comprehensive study on air pollution and green house gases (GHGs) in Delhi. A report submitted to Government of NCT Delhi and DPCC Delhi, 1–334. (accessed 20 December 2020).
Sharma, S.K., Mandal, T.K., Sharma, A., Jain, S. (2018). Carbonaceous species of PM 2.5 in megacity Delhi, India during 2012–2016. Bull. Environ. Contam. Toxicol. 100, 6952.5701. https://doi.org/10.1007/s00128-018-2313-9
Silva, R.A., West, J.J., Lamarque, J.-F., Shindell, D.T., Collins, W.J., Faluvegi, G., Folberth, G.A., Horowitz, L.W., Nagashima, T., Naik, V., Rumbold, S.T., Sudo, K., Takemura, T., Bergmann, D., Cameron-Smith, P., Doherty, R.M., Josse, B., MacKenzie, I.A., Stevenson, D.S., Zeng, G. (2017). Future global mortality from changes in air pollution attributable to climate change. Nature Clim. Change 7, 647–651. https://doi.org/10.1038/nclimate3354
Sreekanth, V., Mahesh, B., Niranjan, K. (2018). Gradients in PM 2.5 over India: Five city study. Urban Clim. 25, 992.5108. https://doi.org/10.1016/j.uclim.2018.06.001
Tyagi, B., Singh, J., Beig, G. (2020). Seasonal progression of surface ozone and NO x concentrations over three tropical stations in North-East India. Environ. Pollut. 258, 113662. https://doi.org/10.1016/j.envpol.2019.113662
Venkataraman, C., Brauer, M., Tibrewal, K., Sadavarte, P., Ma, Q., Cohen, A., Chaliyakunnel, S., Frostad, J., Klimont, Z., Martin, R.V., Millet, D.B., Philip, S., Walker, K., Wang, S. (2018). Source influence on emission pathways and ambient PM 2.5 pollution over India (2015–2050). Atmos. Chem. Phys. 18, 8017–8039. https://doi.org/10.5194/acp-18-8017-2018
Wang, Y., Chen, Y. (2019). Significant climate impact of highly hygroscopic atmospheric aerosols in Delhi, India. Geophys. Res. Lett. 46, 55352.55545. https://doi.org/10.1029/2019GL082339
Yang, X., Jiang, L., Zhao, W., Xiong, Q., Zhao, W., Yan, X. (2018). Comparison of ground-based PM 2.5 and PM 10 concentrations in China, India, and the US. Int. J. Environ. Res. Public Health 15, 1382. https://doi.org/10.3390/ijerph15071382
Zhu, J., Liao, H., Li, J. (2012). Increases in aerosol concentrations over eastern China due to the decadal‐scale weakening of the East Asian summer monsoon. Geophy. Res. Lett. 39, L09809. https://doi.org/10.1029/2012GL051428

Share this article with your colleagues

Research Regions

Middle East
South America
South East Asia

Article

Volume 19, issue 17
ACP, 19, 11159–11183, 2019
Peer review
Related articles

The importance of the representation of air pollution emissions for the modeled distribution and radiative effects of black carbon in the Arctic

Jacob schacht, bernd heinold, johannes quaas, john backman, ribu cherian, andre ehrlich, andreas herber, wan ting katty huang, yutaka kondo, andreas massling, p. r. sinha, bernadett weinzierl, marco zanatta.

Aerosol particles can contribute to the Arctic amplification (AA) by direct and indirect radiative effects. Specifically, black carbon (BC) in the atmosphere, and when deposited on snow and sea ice, has a positive warming effect on the top-of-atmosphere (TOA) radiation balance during the polar day. Current climate models, however, are still struggling to reproduce Arctic aerosol conditions. We present an evaluation study with the global aerosol-climate model ECHAM6.3-HAM2.3 to examine emission-related uncertainties in the BC distribution and the direct radiative effect of BC. The model results are comprehensively compared against the latest ground and airborne aerosol observations for the period 2005–2017, with a focus on BC. Four different setups of air pollution emissions are tested. The simulations in general match well with the observed amount and temporal variability in near-surface BC in the Arctic. Using actual daily instead of fixed biomass burning emissions is crucial for reproducing individual pollution events but has only a small influence on the seasonal cycle of BC. Compared with commonly used fixed anthropogenic emissions for the year 2000, an up-to-date inventory with transient air pollution emissions results in up to a 30 % higher annual BC burden locally. This causes a higher annual mean all-sky net direct radiative effect of BC of over 0.1 W m −2 at the top of the atmosphere over the Arctic region (60–90 ∘ N), being locally more than 0.2 W m −2 over the eastern Arctic Ocean. We estimate BC in the Arctic as leading to an annual net gain of 0.5 W m −2 averaged over the Arctic region but to a local gain of up to 0.8 W m −2 by the direct radiative effect of atmospheric BC plus the effect by the BC-in-snow albedo reduction. Long-range transport is identified as one of the main sources of uncertainties for ECHAM6.3-HAM2.3, leading to an overestimation of BC in atmospheric layers above 500 hPa, especially in summer. This is related to a misrepresentation in wet removal in one identified case at least, which was observed during the ARCTAS (Arctic Research of the Composition of the Troposphere from Aircraft and Satellites) summer aircraft campaign. Overall, the current model version has significantly improved since previous intercomparison studies and now performs better than the multi-model average in the Aerosol Comparisons between Observation and Models (AEROCOM) initiative in terms of the spatial and temporal distribution of Arctic BC.

Article (PDF, 13209 KB)
Article (13209 KB)
Full-text XML

Schacht, J., Heinold, B., Quaas, J., Backman, J., Cherian, R., Ehrlich, A., Herber, A., Huang, W. T. K., Kondo, Y., Massling, A., Sinha, P. R., Weinzierl, B., Zanatta, M., and Tegen, I.: The importance of the representation of air pollution emissions for the modeled distribution and radiative effects of black carbon in the Arctic, Atmos. Chem. Phys., 19, 11159–11183, https://doi.org/10.5194/acp-19-11159-2019, 2019.

The near-surface temperatures in the Arctic are warming at about twice the rate of the global average ( Trenberth et al. , 2007 ; Wendisch et al. , 2017 ). Global climate models have struggled to reproduce the strength of this Arctic-specific enhanced warming, which is commonly referred to as Arctic amplification (AA; Shindell , 2007 ; Sand et al. , 2015 ). Aerosol particles have the potential to substantially affect the Arctic climate by modulating the Arctic energy balance through direct and indirect radiative effects. Considering these effects in models is mandatory for reproducing the observed Arctic amplification ( Shindell et al. , 2009 ) . Within the aerosol population, black carbon (BC) is considered to be the strongest warming short-lived radiative forcing agent ( Quinn et al. , 2015 ) , mainly by absorption of solar radiation in the atmosphere and by reducing the albedo of snow and sea-ice surfaces when deposited. The direct radiative effect of BC on the Arctic has been shown to depend on many factors. Kodros et al. ( 2018 ) show that different assumptions about the mixing state of BC modulate the magnitude of the direct radiative effect, while its sign largely depends on the albedo of the underlying surface. Sand et al. ( 2013 ) come to the conclusion that an increase in BC burdens in the mid-latitudes could have a stronger effect on Arctic sea-ice concentrations and temperatures than an increase in BC concentrations in the Arctic by modulating the meridional energy transport.

The main sources of Arctic BC are located outside of the Arctic circle and originate mainly from fossil fuel use and biomass burning. Local emissions exist in the form of shipping, domestic fuel burning in remote locations, gas flaring and biomass burning ( Stohl et al. , 2013 ) . With declining sea-ice concentrations, the emissions from local shipping are expected to increase ( Corbett et al. , 2010 ; Gilgen et al. , 2018 ). Though human activities in northern Russia represent an important source of BC in the Arctic, these emissions are often underrepresented in recent emission inventories, often missing gas flaring ( Stohl et al. , 2013 ; Huang et al. , 2015 ). Gas flaring is important for the Arctic because of its close vicinity ( Stohl et al. , 2013 ) .

The concentration of BC and other aerosol types like organic carbon, sulfate and dust in the Arctic is the highest in late winter and/or early spring and shows a minimum during the summer. The maximum is often referred to as Arctic haze and is caused by the southward expansion of the Arctic front, which promotes the transport of pollutants from the mid-latitude emission zones ( Law and Stohl , 2007 ) . The Arctic front is a barrier of air with a colder potential temperature, which impedes mixing of air mass, reducing wet removal ( Shaw , 1995 ) . In summer, the northward retreat of the Arctic front, combined with an intensification of precipitation events, leads to a minimum in the aerosol concentration ( Law and Stohl , 2007 ) . Koch et al. ( 2009 ) show that the observed seasonal variability in BC concentrations is challenging for global aerosol models. They showed a tendency to underestimate peak near-surface BC concentrations in late winter and/or early spring ( Shindell et al. , 2008 ) . Although more recent studies show an improvement in the representation of the high late winter and/or early spring concentrations (e.g., Eckhardt et al. , 2015 ; Sand et al. , 2017 ), the model-to-model variability in simulated BC concentration remains considerable ( Eckhardt et al. , 2015 ) .

Despite a good agreement between BC obtained from models and observations close to source regions ( Bond et al. , 2013 ) , in the remote Arctic regions, models still tend to predict too low a BC concentration at the surface in winter and spring, while only some models overestimate it ( Eckhardt et al. , 2015 ) . However, in the upper troposphere, models tend to overestimate the BC concentrations ( Schwarz et al. , 2013 ) . This is caused by a misrepresentation of the aerosol removal processes and transport ( Schwarz et al. , 2013 ) . The mixing and aging, as well as the related removal of aerosol particles along the various transport pathways, are important processes that need to be described accurately in the models ( Vignati et al. , 2010 ) . The representation of emissions is, however, a prerequisite for correctly simulating the transport fluxes and is therefore a key source of uncertainties ( Stohl et al. , 2013 ; Arnold et al. , 2016 ; Winiger et al. , 2017 ). Both the coverage of all BC sources and their temporal variability contribute to the ability to reproduce vertical BC distributions ( Stohl et al. , 2013 ) . The relative source contributions are, however, still discussed with different results ( Winiger et al. , 2017 ) . Bond et al. ( 2004 ) estimate the uncertainty in BC emission inventories to be a factor of about 2. Flanner et al. ( 2007 ) conclude that for the climate forcing by BC in snow, the emissions introduce a bigger uncertainty than the scavenging by snowmelt water and snow aging. However, the quality of emission inventories is difficult to assess with models because of the dependence on the model ( Vignati et al. , 2010 ) .

Having pointed out the potential importance of BC for the AA and the additional uncertainties in aerosol-climate models, in this study, we thoroughly evaluate the global aerosol-climate model ECHAM-HAM for the period 2005 to 2017, with a focus on BC in the Arctic. The evaluation uses a comprehensive set of ground and airborne in situ measurements of BC all across the Arctic and throughout all seasons. In order to address emissions as one of the main sources of uncertainty, we make use of different emission setups to assess the sensitivity of our model to the emission data used. The emissions are composed of different state-of-the-art and widely used emission inventories of anthropogenic air pollution and wildfires. The sensitivity studies allow for estimating the uncertainty range of the BC burden and climate radiative effects in recent aerosol-climate model simulations that are related to emission uncertainties. Estimates of BC radiative effects presented in this study comprise the atmospheric radiative perturbation and the BC-in-snow albedo effect. The model results utilizing the different emission inventories are compared among each other in such a way that the following three points can be explored: (1) the importance of considering daily varying biomass burning emissions, (2) uncertainties in current anthropogenic emission inventories and (3) the potential improvements by regional refinements, in particular in Russian air pollution sources, including gas flaring.

The methods used in this study are discussed in Sect. 2 , with an overview of the model setup and in situ measurements. The sensitivity and related uncertainty in the atmospheric BC burden will be explored in Sect. 3 . Section. 4 will then discuss how well the model performs with the different setups in comparison to BC concentrations obtained by the in situ measurements. Finally, we provide an up-to-date evaluation of the direct radiative effect of BC in the Arctic region and quantify an uncertainty range for this effect that is related to the different emissions (Sect. 5 ).

2.1 Model description

For this study the global aerosol-climate model ECHAM-HAM is used. It was first described in Stier et al. ( 2005 ) . We used the latest version ECHAM6.3-HAM2.3 developed by the HAMMOZ community, ECHAM6-HAM2.3 ( Tegen et al. , 2019 ) . The model is based on the general circulation model ECHAM, developed by the Max Planck Institute for Meteorology (MPI-M) in Hamburg ( Stevens et al. , 2013 ) . ECHAM is coupled online to the aerosol module HAM that is described in detail in Zhang et al. ( 2012 ) . It uses the aerosol microphysics module M7 ( Vignati et al. , 2004 ; Zhang et al. , 2012 ), in which BC, sulfate (SU), organic carbon (OC), sea salt (SS) and mineral dust (DU) are the aerosol species that are accounted for. Volcanic emissions are prescribed. The emission fluxes of mineral dust from deserts as well as sea salt and dimethyl sulfide (DMS) originating from the ocean are calculated online, depending on the meteorology (see Zhang et al. , 2012 ; Tegen et al. , 2019 ). Anthropogenic and biomass burning aerosol emissions are prescribed from emission inventories for which different setups are available.

The aerosol number concentration as well as the mass concentration are prognostic variables calculated using a “pseudomodal” approach. The log-normal modes represent the following: the nucleation mode with a dry radius ( r dry ) range of 0–5 nm and a geometric standard deviation ( σ ln ( r ) ) of 1.59, Aitken mode ( r dry =5 –50 nm, σ ln ( r ) =1.59 ), accumulation mode ( r dry =50 –500 nm, σ ln ( r ) =1.59 ), and coarse mode ( r dry >500 nm, σ ln ( r ) =2.0 ). The latter three exist as hydrophilic and hydrophobic (commonly referred to as soluble and insoluble, respectively). Aerosol in the nucleation mode is always considered hydrophilic, consisting solely of sulfate. The hydrophobic Aitken mode contains BC and OC. In the hydrophilic Aitken mode, they are internally mixed with SU. The hydrophobic accumulation and coarse mode only contain DU. The hydrophilic accumulation and coarse mode contain BC, OC, DU and SS, all internally mixed with SU (see Table 1 ).

The accumulation and coarse modes contain BC, OC and DU, for both classes, and SU (internally mixed), as well as SS, for the mixed classes. Aerosol particles within a mode are assumed to be internally mixed such that each particle can consist of multiple components. Aerosols of different modes are externally mixed, meaning that they coexist in the atmosphere as independent particles. During the mixing, aging and coagulation processes, which are parameterized in M7, aerosol can grow to a bigger mode and can be coated with sulfate to transfer from the hydrophobic to hydrophilic mode. The median radius of the modes can be calculated from the number and mass concentration.

Table 1 Aerosol modes of the species in ECHAM-HAM, including organic carbon (OC), sulfate (SU), mineral dust (DU) and sea salt (SS)

Download Print Version | Download XLSX

The removal process in ECHAM6.3-HAM2.3 is split between sedimentation, dry deposition and wet deposition. The sedimentation process describes the removal by gravitational settling and is applied only to accumulation- and coarse-mode particles. In the model, dry deposition is due to turbulent mixing and affects all but the nucleation mode particles. In the wet deposition scheme, particles are removed as activated aerosol only if the cloud is precipitating. Additionally, below-cloud scavenging is applied. For more details on the removal processes in ECHAM-HAM, see Zhang et al. ( 2012 ) . Monthly and yearly mean values of BC emissions and deposition fluxes computed by ECHAM-HAM for the Arctic are given in Table 2 . Wet deposition accounts for over 90 % of the BC removal and is therefore a crucial impact factor in the Arctic BC burden.

The modeled spatial aerosol distribution affects the climate simulations through interactions with radiation and clouds. A lookup table with precalculated Mie parameters is used to dynamically determine the particle optical properties, considering their size, composition and water content ( Stier et al. , 2005 ; Zhang et al. , 2012 ). The description of cloud microphysics in ECHAM6.3-HAM2.3 is based on the two-moment scheme of Lohmann et al. ( 2008 ) , which allows us to account for the impact of modeled aerosol populations on the number concentrations of cloud condensation nuclei and ice nucleating particles. Particles can collide with droplets and ice particles after they have formed. For further details on the model system, we refer to Stier et al. ( 2005 ) and Zhang et al. ( 2012 ) .

2.2 Emission inventories

While here we focus on BC, the details on the emissions of other aerosol species can be found in Zhang et al. ( 2012 ) . BC is emitted only in the hydrophobic Aitken mode with a median radius of r dry =30 nm and can grow into the bigger modes by aging and coagulation. It can also become hydrophilic by getting coated with sulfate. In this study, we use and compare four different emission setups, which are built from different emission data sets as described in the following.

We use the emissions developed for the Atmospheric Chemistry and Climate Model Intercomparison Project (ACCMIP) as described by van Vuuren et al. ( 2011 ) . The data have a 0.5 ∘ × 0.5 ∘ horizontal resolution and contain anthropogenic and biomass burning emissions that do not differ between years. The ACCMIP emission inventory is available with historic emissions until the year 2000 and for four different development scenarios linked to the Representative Concentrations Pathways (RCPs) for all later years (2000–2100; Lamarque et al. , 2010 ) . The anthropogenic emissions remain constant throughout the year. The biomass burning emissions vary monthly over the course of 1 year but are only scaled by a factor between the years and do not differ in their location. In this study, we only use year 2000 emissions.

The global emission data set created for Evaluating the Climate and Air Quality Impacts of Short-lived Pollutants (ECLIPSE), version 5a, by Klimont et al. ( 2017 ) includes only anthropogenic emissions. The horizontal resolution is 0.5 ∘ × 0.5 ∘ . Historic emissions are available until 2010, and projections of different industrial development scenarios afterwards are linked to the RCPs. Unlike the ACCMIP emission data set, the anthropogenic emissions seasonally vary for the different sectors, and they also include emissions from gas flaring. However, gas flaring emissions from northern Russia have been considered to be difficult to measure and therefore uncertain and possibly too low in current emission inventories ( Stohl et al. , 2013 ) .

To address the importance of local emissions, we use the anthropogenic BC emission data set for Russian BC, described in Huang et al. ( 2015 ) . It is available for the year 2010 and originally comes in a 0.1 ∘ × 0.1 ∘ horizontal resolution but is interpolated here to model resolution of T63 (approximately 1.8 ∘ ). Since the data set is limited to the area of Russia, we combine it with the ECLIPSE emission data. The Russian emissions are distributed between the different months with the monthly patterns of ECLIPSE. The emissions of Russian gas flaring are more than 40 % higher than those in ECLIPSE, resulting partly from a high conversion factor estimated for the fossil fuels found in Russia ( Huang et al. , 2015 ) . It represents a reasonable yet high estimate of local emissions and is used as the reference setup. When compared, the ECLIPSE and Huang et al. ( 2015 ) emissions span an uncertainty range concerning gas flaring emissions.

GFAS (Global Fire Assimilation System) is a data set of biomass burning emissions. The strength of the emissions is scaled to the fire radiative power as observed by the MODIS instruments aboard NASA's Aqua and Terra satellites ( Kaiser et al. , 2012 ) . This allows for a representation of real-time fires in ECHAM-HAM with daily changing emissions and enables it to reproduce the biomass burning plumes, which are regularly observed in the Arctic during spring and summer months. This covers natural fire events as well as those caused by anthropogenic activities. In ECHAM-HAM the biomass burning emissions are injected into the boundary layer regardless of the actual injection height provided by GFAS, which is usually reasonable for most small and moderate boreal fires, while the injection height can be underestimated for specific events with high fire radiative power ( Sofiev et al. , 2009 ) . In previous works with ECHAM-HAM, GFAS emissions were often used with an emission factor of 3.4, as proposed by Kaiser et al. ( 2012 ) . In an early setup, this led to a strong overestimation in BC concentrations in comparison to ground-based and airborne observations at mid-latitudes and high latitudes; it was therefore discarded.

https://www.atmos-chem-phys.net/19/11159/2019/acp-19-11159-2019-f01

Figure 1 Regions indicate the area used for averaging presented in Table 3 . North America is in blue, Europe is in green, Russia is in red and central Asia is in orange.

Table 2 Arctic BC budget averaged for the years 2005–2015 (in kt month −1 ) for BCRUS.

2.3 Experimental setup

We run ECHAM6.3-HAM2.3 at T63 horizontal resolution (approximately 1.8 ∘ ), with 47 vertical layers. The model is driven with ERA-Interim reanalysis data and prescribed sea surface temperature (SST) as well as sea-ice concentrations (SIC). The model simulations cover the 1-year period 2005–2015, with a spin-up period of 3 months. One run is extended to June 2017 in order to include the period of a recent aircraft campaign. In total four model runs are realized, each with a different combination of emission data sets as described in the following. The time-averaged land emissions of BC from each setup are presented for different geographical regions in Table 3 (see Fig. 1 for location).

https://www.atmos-chem-phys.net/19/11159/2019/acp-19-11159-2019-f02

Figure 2 Multi-year monthly mean emissions of (a, b) BC and (c, d) SO X ( SO 2 plus SO 4 ) for the years 2005–2015. Values are integrated over the latitude band between 60 and 90 ∘ N and between 30 and 60 ∘ N.

For the first run we use the historical year 2000 ACCMIP emissions throughout the whole simulation period. Hereafter, this run is referred to as ACCMIP. ACCMIP emission data are still widely used for model experiments, in some cases using the RCPs ( Lund et al. , 2018 ) , or fixed for the year 2000 ( Sand et al. , 2017 ) . This simplification is a common approach to reduce degrees of freedom and control boundary conditions in non-transient climate studies. ACCMIP is the only run that does not use the daily updated GFAS emissions for biomass burning and can therefore not be expected to reproduce actual biomass burning events. Therefore, it can serve as a reference run needed to estimate the uncertainty that is related to the representation of biomass burning emission. The resulting monthly BC for the latitude bands of 30–60 and 60–90 ∘ N can be seen in Fig. 2 . Sulfate is important for the aging and wet removal of BC; therefore the SO 2 plus sulfate ( SO x ) emissions are given as well. It is the run with the highest European emissions, at 538 kt yr −1 , and low central Asian emissions (see Table 3 ). The anthropogenic ACCMIP emissions are higher in Europe than for the other data sets used in this study because recent changes in air quality regulations led to lower emissions there in the period examined (considered until 2011). In Southeast Asia, they are, however, smaller, since the Asian economy has strongly grown since 2000, and with it the air pollutant emissions have also grown.

The second run, called ACCMIP-GFAS, combines the biomass burning emissions of GFAS with the year 2000 ACCMIP emissions from anthropogenic sources (orange line in Fig. 2 ). This run also does not account for changes in anthropogenic emissions but considers the day-to-day variability in wildfires. Together with a setup described in the following, it can be used to assess the range of uncertainty in anthropogenic emissions. This run has the highest average BC emissions in North America, at 515 kt yr −1 , and the lowest central Asian emissions, at 1997 kt yr −1 .

In the third run, we use the ECLIPSE RCP4.5 emission data combined with GFAS emission. It is referred to as ECLIPSE hereafter (blue line in Fig. 2 ). This run has the highest BC emissions in central Asia, with about 1.5 times the emissions of the ACCMIP runs (see Table 3 ).

The fourth run, which is referred to as BCRUS, uses the updated spatially highly resolved BC emissions from Huang et al. ( 2015 ) , replacing and updating only the anthropogenic BC emissions in Russia. Elsewhere the emissions are the same as in the ECLIPSE run. This way, the BC sources are supposed to represent a high estimate, addressing the possibility of underestimation in the global data sets, in particular with respect to gas flaring. For other species, most notably SO X , the runs BCRUS and ECLIPSE do not differ (see green lines in Fig. 2 ).

BCRUS is chosen as the reference run, since it uses the most up-to-date data and is therefore assumed to be the best estimate. In BCRUS the BC emissions north of 60 ∘ N on land are even higher than over the oceans compared with the other data sets, at 172 and 7 kt yr −1 , respectively.

https://www.atmos-chem-phys.net/19/11159/2019/acp-19-11159-2019-f03

Figure 3 Maps of annual mean BC emissions for the years 2005–2015. (a) Absolute values are given for BCRUS. Difference between (b) the ACCMIP-GFAS and BCRUS results, (c) the ECLIPSE and BCRUS results, and (d) between the ACCMIP and ACCMIP-GFAS results.

Figure 3 a shows the emissions of BC for the reference run BCRUS. The highest emissions north of 30 ∘ N are found in the industrial regions of East Asia, Europe and northeastern America as well as in gas and oil extraction areas in North America and northern Russia. The anthropogenic emissions in the sparsely populated northern Canadian and Alaskan regions are much lower than those of the densely populated European region. Additionally, the aforementioned gas flaring emissions in Russia are assumed to be higher than in northern North America. The transport efficiency from the East Asian sources to the Arctic is comparably low, but the high emissions in this region makes it important for long-range transport to the Arctic upper troposphere ( Ikeda et al. , 2017 ) .

Figure 3 b and c show the difference in BC emissions for the ACCMIP-GFAS and ECLIPSE runs compared with the BCRUS setup, respectively. BC emissions from ACCMIP-GFAS are higher than those of BCRUS in North America, Europe, western Russia and Japan. They are, however, lower in northern Russia by more than 3500 kg km - 2 yr - 1 and China by more than 2800 kg km - 2 yr - 1 . In northern Russia and China, however, ACCMIP-GFAS emissions of BC are locally over 3500 and 2800 kg km - 2 yr - 1 lower, respectively. Figure 3 c shows the difference between ECLIPSE and BCRUS. There are only differences in Russia, as expected. The ECLIPSE emissions are smaller than the BCRUS emissions because of newer information about additional sources. Among other sources, higher values are mainly due to gas flaring emissions. Figure 3 d shows the difference in BC emissions between the runs ACCMIP and ACCMIP-GFAS that results from their difference in the biomass burning representation discussed above. ACCMIP shows higher emissions in Europe and Russia, while the emissions of ACCMIP-GFAS are higher in North America. The totals of BC emissions are summarized in Table 3 .

2.4 Calculation of direct aerosol radiative effects of BC

For diagnostic output, the instantaneous radiative impact of all aerosol types is calculated in ECHAM-HAM by calling the radiation routine twice: once considering the interaction between aerosol particles and radiation and once without any aerosol. The difference between these two calls is then considered to be the direct aerosol radiative effect (DRE), which is free of any rapid adjustment (semi-direct effects).

To calculate the DRE by BC, the ACCMIP-GFAS and BCRUS runs were repeated, leaving BC out in the computation of radiative fluxes. For this, BC was skipped in the calculation of the complex radiative index and the radiatively active number of particles, while the wet radius of respective aerosol modes was not adjusted further. The DRE of BC is then derived from the difference of these two runs to their original setup. Note that with this method, the estimate includes the semi-direct effect of BC, which is small in the large-scale average, since positive and negative effects cancel each other out, and is not statically significant in the Arctic ( Tegen and Heinold , 2018 ) . The DRE of BC is studied for the sub-period 2005–2009.

The aerosol transport and radiation simulations in this study consider the reduction of snow albedo due to deposited BC. The BC-in-snow albedo effect is parameterized in terms of a lookup table based on a single-layer version of the Snow, Ice and Aerosol Radiation (SNICAR) model from Flanner et al. ( 2007 ) . The scheme was first implemented in the earth-system model version of ECHAM6 by Engels (2016) and has become available recently in ECHAM6.3-HAM2.3 ( Gilgen et al. , 2018 ) . It accounts for the BC concentration within in the uppermost 2 cm of snow. Input parameters are the snow precipitation, the sedimentation, dry deposition and wet removal of BC as well as on snowmelt and glacier runoff, the latter of which leading to an enrichment of BC in the remaining snow layer. The BC-in-snow albedo effect is computed for solar radiation only because the albedo is only used for shortwave (solar near-infrared and visible) wavelengths in the model. The effect in the terrestrial spectrum is very small and can be neglected for the atmosphere. A feature not considered so far is the impact of BC deposition on bare sea ice. This, however, is expected to be negligible, since the spatial extent of sea ice without snow cover is small ( Gilgen et al. , 2018 ) . In this study, the parameterization is only used for diagnostics of the BC-in-snow albedo effect without any feedback on the model dynamics.

https://www.atmos-chem-phys.net/19/11159/2019/acp-19-11159-2019-f04

Figure 4 Geographic regions of Arctic aircraft campaigns. The data of these are used for model evaluation: HIPPO in blue, ACLOUD and PAMARCMiP-2017 in green, ACCESS in red, and ARCTAS in orange. Black triangles show the location of stations with BC surface measurements.

Table 3 Area-weighted totals of BC emissions from anthropogenic sources and biomass burning fires for the main source regions (as shown in Fig. 1 ) averaged for the years 2005–2015 (in kt yr −1 ).

2.5 Observations

2.5.1 near-surface bc concentrations.

Near-surface BC concentrations are taken from different measurement sites around the Arctic, shown on the map in Fig. 4 as triangles. These sites utilize different measurement principles providing BC concentrations that differ by definition. The measurement principles, measurement period and the location of the measurements are summarized in Table 4 .

At five of these sites light absorption was measured with aethalometers. Of those five stations, Alert and Summit measured at 467, 525 and 637 nm; Zeppelin Station measured at 525 nm only; and Tiksi and Pallas measured at 637 nm only. From the light absorption, equivalent black carbon (eBC) concentrations were calculated using different mass absorption coefficients (MACs) depending on the wavelengths. For the stations where measurements at 525 nm were available, 9.8 m 2 g −1 was used for aged Arctic BC at 550 nm (from Zanatta et al. , 2018 ) . Zanatta et al. ( 2018 ) give an uncertainty of the MAC value of ±1.68 m 2 g −1 . This implies an uncertainty range of approximately −20 % to +15 % for the observed BC concentrations. For the stations where the light absorption was only available at 637 nm, a MAC of 8.5 m 2 g −1 for Scandinavian BC was used at this wavelength correspondingly ( Zanatta et al. , 2016 ) . The data were processed as described in Backman et al. ( 2017 a ) to reduce noise and lower the detection limit, which is important for the Arctic, since concentrations tend to be about 1 order of magnitude lower than at mid-latitudes outside of Arctic haze season. We use the variable collection time data from Backman et al. ( 2017 b ) , which covers January 2012 until December 2014. The sites are Alert in Nunavut, Canada; Pallas in Finland; Tiksi in Sakha, Russia; Zeppelin Station in Svalbard, Norway; and Summit in Greenland, Denmark.

BC concentration data measured with a continuous soot-monitoring system (COSMOS), which removes volatile aerosol compounds, are available for Ny-Ålesund, Svalbard, Norway, and Barrow, Alaska, USA, for the period from 1 April 2012 to 31 December 2015 and 12 August 2012 to 31 December 2015, respectively. The data collection and the retrieval of BC mass concentrations using a MAC of 8.73 m 2 g −1 at 565 nm are described in Sinha et al. ( 2017 ) .

In addition, we use measurements of eBC concentration at the Villum Research Station in northern Greenland that were performed with a multi-angle absorption photometer (MAAP). We use daily averaged data from 14 May 2011 to 23 August 2013. Further information on data sampling and processing can be found in Massling et al. ( 2015 ) .

For Alaska we use filter-collected BC data acquired by the Interagency Monitoring of Protected Visual Environments (IMPROVE) aerosol network. The thermal protocol used to process the measurements is described in Chow et al. ( 2007 ) . We use data from the sites Tuxedni, Trapper Creek, Denali National Park (NP) and Gates of the Arctic NP.

2.5.2 Aircraft campaigns

The correct representation of the modeled aerosol vertical distribution is a key prerequisite for estimating the aerosol radiative impact ( Samset et al. , 2013 ) . For this reason, we collected BC measurements from five Arctic airborne campaigns. During all flights, the mass concentration of refractory black carbon (rBC) was quantified by means of the single particle soot photometer (SP2), which ensures the high time resolution and high sensitivity required in airborne observations.

The HIPPO (HIAPER Pole-to-Pole Observation) campaign consists of five deployments by the National Science Foundation (NSF; data set – Wofsy et al. , 2017 , version 1): HIPPO-1 (9 to 23 January 2009), HIPPO-2 (31 October to 22 November 2009), HIPPO-3 (24 March to 16 April 2010), HIPPO-4 (14 June to 11 July 2011) and HIPPO-5 (9 August to 8 September 2011). Flights included Northern Hemisphere high latitudes over North America, the Pacific Ocean and the Bering Sea. BC particles were measured with an SP2. The aircraft used was the NSF/National Center for Atmospheric Research (NCAR) Gulfstream-V (GV).

The BC data from NASA's campaign ARCTAS (Arctic Research of the Composition of the Troposphere from Aircraft and Satellites) were collected in two deployments, spring (April 2008) and summer (June–July 2008), over North America and the American Arctic. The mission design and execution are described in Jacob et al. ( 2010 ) (data set – SP2_DC8; https://www-air.larc.nasa.gov/cgi-bin/ArcView/arctas/ , last access: 2 July 2018).

The summer campaign of ACCESS (Arctic Climate Change, Economy and Society) in July 2012 took place over Scandinavia and the European Arctic ( Roiger et al. , 2015 ) . The BC mass concentration was derived from measurements of a SP2 aboard the Falcon aircraft of the DLR (Deutsches Zentrum für Luft und Raumfahrt).

Another set of airborne measurements was collected from the 2017 PAMARCMiP (Polar Airborne Measurements and Arctic Regional Climate Model Simulation Project) campaign ( Herber et al. , 2012 ) . The selected flight took place in March and was based in Longyearbyen, Spitzbergen, Norway, and made use of the Polar 5 aircraft of the Alfred Wegener Institute (AWI).

Also based in Ny-Ålesund was the ACLOUD (Arctic CLoud and Observations Using airborne measurements during polar Day) campaign, with measurements from 22 May to 28 June 2017 ( Wendisch et al. , 2018 ) . Again, the BC concentrations were measured with an SP2 aboard the AWI Polar 5 aircraft.

The range of flight tracks of the aircraft campaigns used in this study are mapped in Fig. 4 as colored boxes, with HIPPO in blue, ACCESS in red, ARCTAS in orange, and the combination of ACLOUD and PAMARCMiP-2017 in green. The most western, eastern, southern and northern coordinates at which the aircraft took measurements form the edges of the boxes, with measurements south of 60 ∘ N not being considered. An overview of instruments and dates is given in Table 4 . Even though aircraft campaigns can only give information within a short time window, the combination of different campaigns allows us to cover the almost entire year except for December, February, September and October, the months for which no aircraft data are available.

The comparison between a coarsely resolved model and aircraft measurements is challenging because of many factors. Any observed feature of subscale lifetime or spatial extend will be missed or at least underestimated by a model that is designed to estimate climate-relevant effects over multiple years. Schutgens et al. ( 2016 ) suggest either spatio-temporal averaging of both measurements and spacial interpolated model data or increasing the model resolutions to achieve the best agreement. Lund et al. ( 2018 ) show that using only monthly mean model output introduces significant biases.

In this study, we sample from the model's 12-hourly output for each measurement point during one campaign before averaging to one vertical profile, without prior interpolation.

In order to investigate the uncertainty range in the BC burden and its direct radiative impact, which results from the uncertainty in emissions, different simulations with the aerosol-climate model ECHAM6.3-HAM2.3 using four emission configurations are performed and compared as outlined in Sect. 2.3 .

https://www.atmos-chem-phys.net/19/11159/2019/acp-19-11159-2019-f05

Figure 5 Contour plot showing the modeled atmospheric BC burden averaged over the simulation period (2005–2015). (a) Absolute values from the BCRUS setup, which is used as the reference. (b, c) Differences of ACCMIP-GFAS and ECLIPSE to the BCRUS run, respectively. Blue colors indicate lower BC burden than in the BCRUS run, and red indicates higher BC burden. (d) Difference in modeled atmospheric BC burden between ACCMIP and ACCMIP-GFAS.

The atmospheric burden of BC averaged over the simulation period (2005–2015), which results from the different emission setups, is shown in Fig. 5 . The distribution over the BC burden resulting from BCRUS (see Fig. 5 a) is comparable to the distribution of the emissions in this run (see Fig. 3 a). The northward transport results in a visible separation between the Eastern and Western Hemispheres in the BC burden in the northern part of 60 ∘ N, with higher values of 200 to 800 µ g m −2 in the Eastern Hemisphere compared with values of 50 to 400 µ g m −2 in the Western Hemisphere. This separation along the prime meridian is a result of higher anthropogenic emissions in the north of the Eastern Hemisphere, as discussed in Sect. 2.3 . The area-weighted mean burden of BC north of 60 ∘ N of BCRUS is 254 µ g m −2 in the multi-year annual average, which is the highest among the model runs used for this study. The highest values north of 60 ∘ N are located in the Russian gas flaring region, at over 560 µ g m −2 .

The causes and details, as well as differences between the runs, will be discussed in the following.

3.1 Recent economic changes

To estimate the range of anthropogenic emissions in currently widely used inventories, we compare the runs BCRUS and ACCMIP-GFAS. The ACCMIP run does not take recent economic changes into account, since emissions are fixed to the year 2000. BCRUS, on the other hand, is largely based on the ECLIPSE emissions that consider the economic development until 2015 and provide projections for the years after. Since both are combined with the biomass burning emissions from GFAS (which covers natural as well as human-caused fires), the differences in BC emissions are solely in the anthropogenic emissions (excluding human-caused grass and forest fires).

The use of fixed emissions in ACCMIP-GFAS causes a remarkable difference in the atmospheric burden of BC over the source regions compared with the reference run (see Fig. 5 b). ACCMIP-GFAS does not catch the reduction in BC emissions over western countries and Japan due to the implementation of strict air quality legislation and the increased emission over China caused by its economic growth. The neglect of the recent economic evolution and mitigation policies results in an overall underestimation of the BC burden by 63 µ g m −2 (25 %) within the 60–90 ∘ N latitudinal band. Over the Kara Sea, the result is an underestimation that exceeds 100 µ g m −2 ; this is a region that has been considered to be a hotspot for the connection between Arctic sea-ice loss and changes in the large-scale atmospheric circulation with particular sensitivity (e.g., Petoukhov and Semenov , 2010 ).

3.2 Regional refinement

Higher, more realistic estimates of emissions for Arctic sources (e.g., gas flaring) have been discussed as a requirement for reproducing observations like locally high BC concentrations in snow ( Eckhardt et al. , 2017 ) as well as the layering and seasonality of Arctic aerosol concentration far from sources ( Stohl et al. , 2013 ) . However, improving the regional accuracy of BC emissions in the Russian Arctic does not impact the modeled BC spatial distribution meaningfully outside of the Russian Arctic. As seen in the comparison of the runs BCRUS and ECLIPSE (Fig. 5 c), the difference in the BC burden between BCRUS and ECLIPSE shows only differences visible in Russia, since the BC emissions differ only there (see Table 3 and Fig. 3 ). This results in an increase in the BC burden, mainly in the eastern Arctic, with up to 25 µ g m −2 higher values over the Barents Sea and Kara Sea. The area-weighted annual averages north of 60 ∘ N differs by 11 µ g m −2 , with the higher BC burden being produced by BCRUS. However, stronger effects are found for BC near-surface concentrations as discussed below, due to the vicinity of the refined sources to the Arctic and the resulting transport at the lowest atmospheric levels.

3.3 Temporal variability in wildfire events

The atmospheric composition and, in particular, the BC loading are strongly influenced by wildfires, which have a strong spatio-temporal variability. The importance of considering actual biomass burning events is demonstrated by comparing the runs ACCMIP-GFAS and ACCMIP. While ACCMIP-GFAS accounts for real fire events derived from satellite retrievals, ACCMIP uses fixed fire emissions for the year 2000. The ACCMIP-GFAS BC emissions are higher than the ones of ACCMIP by 64.5 kt yr −1 ; this is mainly caused by North American emissions (see Fig. 3 d).

The patterns of the BC burden of both runs are similar, with a higher burden over the western industrialized countries and a lower burden over China compared to BCRUS. The area-weighted average burden of BC estimated with ACCMIP is 186 µ g m −2 , which is 11 µ g m −2 (6 %) less than ACCMIP-GFAS. A map of the differences in the annual average burden of BC due to the different representations of biomass burning emissions is shown in Fig. 5 d. A clear pattern of a lower BC burden over southern Siberia and a higher burden over North America is visible. For the high Arctic, both runs produce a similar burden in the 11-year mean, with differences in the BC burden of less than 25 µ g m −2 . However, for short time periods, influenced by biomass burning events, the difference between the two runs can be dramatic, as shown below for comparisons of the BC mass concentration.

4.1 Near-surface BC mass concentration

Near-surface measurements of BC mass concentrations can help evaluate the capability of ECHAM-HAM to reproduce the distribution of BC in the Arctic atmosphere and hence reasonable estimates of the warming influence of absorbing aerosol. While the data are only representative of the lowest atmospheric layer, the long time series give robust information about this specific important climate forcer. The multi-year seasonality of near-surface BC is compared with observations in the Arctic, as is the temporal correlation, with a spatial emphasis. Each measurement point is compared with the nearest grid cell at the closest time step from the model. The medians are calculated after this sampling.

https://www.atmos-chem-phys.net/19/11159/2019/acp-19-11159-2019-f06

Figure 6 Near-surface BC mass concentrations for Atlantic Arctic stations. Solid black line shows the multi-year monthly median BC mass concentration observed in (a) Zeppelin Station, (b) Ny-Ålesund, (c) Villum and (d) Alert. See Fig. 9 for the geographical locations. Dashed black line indicates the observed upper and lower quartiles. In color are the median different model runs with solid lines and filled circles, and the upper and lower quartiles run with empty circles.

https://www.atmos-chem-phys.net/19/11159/2019/acp-19-11159-2019-f07

Figure 7 As in Fig. 6 for the stations in (a) Tiksi, (b) Pallas, (c) Barrow and (d) Summit.

Figures 6 through 8 each show the comparison of the observed and modeled monthly median mass concentration of near-surface BC for four available Arctic field sites averaged over multiple years. A list with detailed information on measurement period, instrumentation and data providers can be found in Table 4 . The model is compared with the measurements in terms of how well the annual cycle is reproduced by comparing median BC mass concentration values and in terms of the ability to reproduce pollution events at the correct time by analyzing the Pearson correlation coefficients.

https://www.atmos-chem-phys.net/19/11159/2019/acp-19-11159-2019-f08

Figure 8 As in Fig. 6 for the American Arctic stations of the IMPROVE network in Gates of the Arctic NP (a) , Trapper Creek (b) , Tuxedni (c) and Simeonof (d) .

Of the stations used, Zeppelin Station and Ny-Ålesund are located in Svalbard. Alert and the Villum Research Station are both situated in the north of the Greenland ice sheet. The annual cycle of the BC concentration is shown in Fig. 6 in terms of the median, upper and lower quartiles in black; the different model runs are color coded. At all four stations, the maximum median BC mass concentration is observed in spring, at 36, 72 and 73 ng m −3 for Zeppelin Station, the Villum Research Station and Alert in March, respectively.

For Ny-Ålesund the highest concentrations are observed in April, with a median of 30 ng m −3 . For all stations in Fig. 6 there is a minimum in summer, with less than 15 ng m −3 median BC concentrations in the near-surface air. At all four stations, the reference run BCRUS produces higher median concentrations in January than observed. The modeled BC mass concentrations are underestimated by the model at all of these stations except Ny-Ålesund, at least for some months. The model overestimates the BC concentrations in the beginning of the year at all stations. The overestimation is largest at Ny-Ålesund, with monthly median values of up to 120 ng m −3 for BCRUS, compared with the measured median of 20 ng m −3 .

For Zeppelin Station and Ny-Ålesund, BC is also overestimated in November and December. Here, the model simulates monthly median values, each at 90 ng m −3 , for December compared with measured medians of 10 and 20 ng m −3 for Zeppelin Station and Ny-Ålesund, respectively. It has to be noted that, in the model resolution, Zeppelin Station and Ny-Ålesund are in the same grid box. The difference in altitude is not taken into account from the model side; instead the lowest level above the modeled orography is chosen. Differences in the model results between the two stations, shown in Fig. 6 , are only due to the different temporal availability of the measurements. Interestingly the model agrees slightly better with the observations at Zeppelin Station, which is more exposed to long-range transport, while Ny-Ålesund is often subject to a blocking situation that prevents mixing of air mass because of its respective location.

Figure 7 shows the second set of stations. Tiksi, Pallas and Utqiaġvik (Barrow) show the same annual cycle as the stations in Fig. 6 , with high concentrations in winter and spring as well as minimum concentrations in summer. The model slightly underestimates BC at Tiksi in all months, with high concentrations of over 50 ng m −3 . For Pallas and Utqiaġvik (Barrow) an overestimation by the model is found for January. Summit shows a different annual cycle in the observations, with the highest median BC mass concentrations of slightly more than 30 ng m −3 being observed in April and with slightly lower values in summer, and a minimum was observed for January. The model was neither able to reproduce this different annual cycle nor the peak in the quartiles during September and December that were observed. However, the amount of BC mass agrees well between model and measurements, with values generally below 30 ng m −3 .

Results for four Alaskan stations of the IMPROVE network are shown in Fig. 8 . There, the highest median BC concentrations are observed in the summer months, at 70 ng m −3 at Gates of the Arctic NP in June, 50 ng m −3 at Trapper Creek in July, and 40 and 60 ng m −3 at Tuxedni and Simeonof in August, respectively. The model noticeably fails to reproduce these summer maxima and instead produces the highest concentrations in January to March in a similar way as for the other Arctic stations. In Tuxedni, the simulated median concentration lies at 60 ng m −3 , while the observed one is at 10 ng m −3 . The Brooks Range spans through Alaska, from the Bering Sea in the west to the Beaufort Sea in the east, with multiple peaks of more than 2000 m a.s.l. Situated south of Brooks Range, the four stations are shielded from the Arctic. An underestimation of the orographic height in the coarsely resolved model could therefore be the reason for this misrepresentation.

https://www.atmos-chem-phys.net/19/11159/2019/acp-19-11159-2019-f09

Figure 9 Map showing the Arctic sites where the near-surface BC mass concentration was measured. Colors show the correlation coefficient between the measured and modeled daily averages. Correlation coefficients close to zero are not colored. Top right segment indicates the correlation coefficient for the BCRUS run. Clockwise are the ACCMIP, ACCMIP-GFAS and ECLIPSE runs. The label of Zeppelin Station is shifted to the north on the map for better visibility. The label of station Trapper Creek is shifted to the southeast.

The Pearson correlation coefficient between the collocated data of measured and modeled BC mass concentrations for all available aerosol stations in the Arctic region is shown in Fig. 9 . Since pollution events in the Arctic can raise the BC concentrations to levels well above the background, the correlation coefficient is very sensitive to the model being able to reproduce the timing of pollution events. Therefore, this analysis complements the median and quartiles discussed above.

The top right segment of each circle shows the correlation coefficient between the BCRUS model run and the measurements. Following clockwise are the correlations for the runs ACCMIP, ACCMIP-GFAS and ECLIPSE. The circle for Summit is not filled, since there the correlation coefficients are negative albeit close to zero ( −0.06 for BCRUS). The negative correlation corresponds to the opposite annual cycle of surface BC in the BCRUS model results compared with the observations as shown in Fig. 7 . For all other stations, correlation coefficients are positive. Simeonof, on the Alaska Peninsula, shows a very weak correlation, with 0.09 for the different model runs. Tuxedni, on the southern coast of Alaska, also has a relatively low correlation coefficient of 0.44.

For the other Alaskan stations of the IMPROVE network, however, a correlation between observations and BCRUS model results is found that is robustly positive. Even for the stations where the annual cycle was not reproduced, the correct timing of short-term events leads to these positive correlation coefficients. Trapper Creek shows a correlation coefficient of 0.55, Denali NP of 0.72 and Gates of the Arctic NP of 0.94. ACCMIP clearly performs the worst of all experiments, with correlation coefficients 0.14, 0.31 and 0.20 for Trapper Creek, Denali NP and Gates of the Arctic NP, respectively, while the other runs do not differ strongly from each other. Taking the position and strength of actual biomass burning events into account is crucial for correctly reproducing the near-surface BC concentrations in Alaska.

The correlation coefficient at Oulanka is below 0.3 for all runs. This, however, is computed only on the basis of 3 months of measurements. The other European stations of Pallas, Ny-Ålesund and Zeppelin Station also show relatively low correlation coefficients of 0.45, 0.50 and 0.30 for BCRUS, respectively. The other runs behave similarly.

At the four northernmost stations, Tiksi, Utqiaġvik (Barrow), Alert and Villum Research Station, correlation coefficients of 0.55, 0.65, 0.60 and 0.60 are found for BCRUS, respectively. These four stations are located north of a big land mass and likely show a good correlation, since concentrations are drastically different when the wind either comes from the land or the Arctic Ocean. With the exception of Tiksi, the ACCMIP run does not produce considerably weaker correlations with the observations than the other runs. At Tiksi, the highest correlation coefficient is expected for BCRUS, since BCRUS comprises the most recent and detailed emissions specifically for Russia. At 0.56 compared with 0.71 (ACCMIP-GFAS) and 0.61 (ECLIPSE), the correlation is, however, the lowest.

4.2 Vertical distribution of BC

The BC mass mixing ratio from airborne measurements is a valuable source of information about the vertical distribution of BC. However, because of the logistical difficulties and high costs, the spatial and temporal coverage is quite sparse. The aircraft campaigns used in this study for model evaluation are described in detail in Sect. 2.5.2 , their geographical operation area is presented in Fig. 4 and they are listed in Table 4 . Each measurement point is compared with the nearest grid cell from the model, resulting in one average profile per campaign and run. We group the campaigns based on season, resulting in at least one profile per season, with better coverage during spring and summer, which have three and four campaigns, respectively.

4.2.1 Winter

For the winter months (December–January–February; DJF) only data from the HIPPO campaign are available, starting with the first deployment during January 2009. We consider only data points north of 60 ∘ N. The area covered by HIPPO is indicated by the blue box in Fig. 4 . As shown in Fig. 10 , observed BC mass mixing ratios were highest near the ground. Everything below 950 hPa is removed from the plot because of unrealistically high measured BC mass mixing ratios near the ground of over 450 ng kg −1 on average, which could not be reproduced by the model. Starting at 950 hPa the simulated profile of BCRUS is very similar to the observed vertical distribution. Model results and measurements show a decrease in BC with height, with the BCRUS run overestimating the BC mass mixing ratio above 900 hPa by a factor of about 2. The ECLIPSE run produces almost the same profile; however the runs ACCMIP and ACCMIP-GFAS produce lower values that, while still higher, are closer to the observed profile. Since the emission of BC for these runs is only lower in central Asia (see Table 3 ), this likely points toward an overestimation of the modeled transport to the Arctic, possibly caused by an underestimation of wet removal.

4.2.2 Spring

The observed and modeled profiles of BC mass mixing ratios from the ARCTAS spring campaign over the American Arctic (orange box in Fig. 4 ) in April 2008 can be found in Fig. 11 a. Observations show high values near the ground, with a BC mass mixing ratio of over 40 ng kg −1 , and a steep increase from there towards a pollution layer, with a maximum of almost 200 ng kg −1 at a 600 hPa height. BCRUS (in green) correctly places this layer but underestimates its strength. A second BC layer is centered at about a 400 hPa height, with the mixing ratio gradually decreasing above. All model runs including actual fire emissions are well able to capture the placement of the aerosol layer, while the magnitude is underestimated by a factor of up to 3. This could be caused by emissions that are too low in the source region with a correctly predicted transport or could just be an effect of the coarse resolution of the model, resulting in the emissions for the fire event being mixed over the grid boxes instead of being concentrated in a confined plume. In particular, small local fire plumes may be too strongly diluted when emitted into the model boundary layer. In addition, there is the possibility of a large sampling bias, with fire plumes being specifically probed during the campaign ( Jacob et al. , 2010 ) . The other runs using GFAS produce similar results, with ECLIPSE and BCRUS performing the best. The ACCMIP run without daily fire emissions deviates most from the observations. This shows that this BC distribution was in fact largely caused by a biomass burning fire plume.

The averaged profile of the measured BC mass mixing ratio for the HIPPO-3 campaign over the Pacific in March–April 2010 is plotted in Fig. 11 b. It shows observed mixing ratios of 20 ng kg −1 near the surface. There is a local minimum at a height of 880 hPa. The highest mass mixing ratio of BC is found at heights around 520 hPa. ECHAM-HAM is able to reproduce this profile well up to a height of about 650 hPa in all runs. From there the model underestimates the amount of BC up to the height of about 400 hPa. Above this, the model overestimates the amount of BC. The overestimation at uppermost levels is twice as high in the ECLIPSE and BCRUS model runs. They likely overestimate the long-range transport from Southeast Asian or Russian pollution sources. Close to the ground, BCRUS and ECLIPSE are better able to reproduce the observed mass mixing ratio.

The ACLOUD campaign took place around Svalbard in May and June 2017 and therefore represents late spring and early summer. As can be seen in Fig. 11 c, the mixing ratio during the ACLOUD campaign was low, with observed mass mixing ratios of 4 to 5 ng kg −1 near the ground. A maximum with 14 ng kg −1 was observed at 800 hPa, above which the mass mixing ratio declined with increasing altitude. ECHAM-HAM reproduced this averaged profile relatively well, only placing the maximum too high at a height of 650 hPa, where the observations again decreased to 4 ng kg −1 . This overshooting by ECHAM-HAM, at upper levels, is mainly found for the last flight on 16 June 2017 (not shown separately). This already hints to the tendency of ECHAM-HAM to overestimate upper-layer transport of aerosol in summer, as described in the text below. Note that for ACLOUD, only BCRUS results can be presented because of the timeliness of the measurements.

https://www.atmos-chem-phys.net/19/11159/2019/acp-19-11159-2019-f10

Figure 10 Vertical profiles of BC mass mixing ratios from airborne in situ measurements during the flight campaign HIPPO-1 campaign in January 2009. The modeled BC mass mixing ratios were averaged over the vertical levels. The observations are shown in black, and the different model runs are color coded (see Sect. 4 for details).

4.2.3 Summer

Results for the comparison between the ARCTAS summer campaign over the American Arctic in June and July 2008 and the model results from ECHAM-HAM are shown in Fig. 12 a. The averaged profile over the campaign shows an increase in the BC mass mixing ratio, with increasing height up to a maximum of 26 ng kg −1 at the 300 hPa level. As discussed by Matsui et al. ( 2011 ) , air mass during this campaign was influenced by biomass burning in eastern Russia. Most of the BC from these fires, however, was quickly removed from the atmosphere by wet depositions by heavy rain close to the source region ( Matsui et al. , 2011 ) . BCRUS produced a similar profile, with BC mass mixing ratio values very close to the observations up to 700 hPa height. Above this level, the model overestimates the amount of BC. This points toward a misrepresentation in the wet removal process or possibly vertical mixing or uplift of fire aerosol that is too efficient in the model. ACCMIP strongly differs from the other runs and observations, producing much higher mixing ratios below 550 hPa height. Above 570 hPa the BC mass mixing ratios modeled by ACCMIP, however, are much closer to the observations. At this height, Matsui et al. ( 2011 ) found elevated values in measured CO, pointing toward an influence by biomass burning fires. ACCMIP agrees best with the measurements because the observed fires that lead to the overestimation in the other runs are not present in the run. In this way, it produces values that are similar to the observations where biomass burning aerosol was removed.

https://www.atmos-chem-phys.net/19/11159/2019/acp-19-11159-2019-f11

Figure 11 As in Fig. 10 , but with spring campaigns (March–April–May). (a) ARCTAS spring campaign in April 2008. (b) HIPPO-3 campaign in March–April 2010. (c) ACLOUD campaign in May–June 2017. Note that for year 2017, model results are only available from the BCRUS run.

Observations from HIPPO-4 (June–July 2011) and model results are compared in Fig. 12 b. Modeled and observed mixing ratios are relatively low, with the highest observed BC mass mixing ratio at just above 19 ng kg −1 . In BCRUS this maximum is found at 820 hPa; this is much lower than observed (620 hPa). The modeled vertical extent of this pollution layer is also thinner than observed. The major difference is BC amounts that are far too high between 500 and 200 hPa in the model results for all emission setups. Noteworthy is also the difference between the runs of ACCMIP and ACCMIP-GFAS, with ACCMIP performing better than the others runs in reproducing the pollution layer in the lower troposphere. The emissions from the actual fires in the GFAS emissions seem to not have reached the observed height but instead mostly remained below 800 hPa. The ACCMIP biomass burning emission coincidentally allowed ECHAM-HAM to reproduce a layer that is influenced by biomass burning in the same height as observed. The fact that all runs that use GFAS produce the same profile, while the only run without it produces a different profile, shows that the BC profile, at least up to a height of 300 hPa, is mainly caused by fire emissions.

https://www.atmos-chem-phys.net/19/11159/2019/acp-19-11159-2019-f12

Figure 12 As in Fig. 10 , but for summer campaigns (June–July–August). (a) ARCTAS summer campaign in June–July 2008. (b) HIPPO-4 campaign in June–July 2011. (c) HIPPO-5 campaign in August–September 2011. (d) ACCESS campaign in June 2012.

The profile plot for HIPPO-5 (August–September 2011) shows low observed and modeled mass mixing ratios throughout the atmosphere (see Fig. 12 c). The observations show the highest mass mixing ratio close to the surface at 9 ng kg −1 and a decrease towards 870 hPa to values just over 1 ng kg −1 . The observed BC mass mixing ratio stays low at layers above. BCRUS produces lower BC mass mixing ratios near the surface and overestimates the amount of BC above 850 hPa. ACCMIP is the only run producing significantly different BC mass mixing ratios from the other runs, with strong overestimation throughout the profile and BC mixing ratios of up to 34 ng kg −1 at a height of 930 hPa. This strong overestimation is related to inappropriate biomass burning emissions in ACCMIP in this area.

https://www.atmos-chem-phys.net/19/11159/2019/acp-19-11159-2019-f13

Figure 13 As in Fig. 10 , but for the fall campaign HIPPO-2 in November 2009.

Figure 12 d shows the BC mass mixing ratio of the ACCESS campaign in June 2012 averaged over all flights, with the exception of the transfer flights. The observations show a decrease from the near-surface mixing ratios of 13 ng kg −1 to a layer of cleaner air at 870 hPa (5 ng kg −1 ). The modeled BC profiles show increasing mass mixing ratios with increasing altitude, with the exception of very high mixing ratios near the ground. The minimum mixing ratios are found at 900 hPa. The model shows a considerable overestimation between 800 and 400 hPa.

4.2.4 Fall

The second mission of the HIPPO campaign measured BC layering over the Pacific during November 2009. The fall profile is shown in Fig. 13 . The highest BC mass mixing ratio of up to 40 ng kg −1 was found near the surface, with a steep decrease to 5 ng kg −1 just below 900 hPa. Above that height there is a lofted BC layer around 420 hPa containing 26 ng kg −1 . The BCRUS run underestimates the mixing ratios at the surface by 14 ng kg −1 . The lofted BC layer is placed slightly too low between 850 and 470 hPa. The amount of BC, however, is well matched. With increasing altitude, the increases in the amount of BC are steeper than in the observations. The other runs show a very similar vertical layering of the modeled BC mixing ratio. ACCMIP and ACCMIP-GFAS underestimate the pollution layer below 500 hPa. Again, the BC mixing ratios are strongly overestimated above 280 hPa, in particular in the runs ECLIPSE and BCRUS. This is either due to an overestimation in the upper-level, long-range transport of North American or Russian air pollution or to an underestimation in removal which could contribute to the upper-level transport.

Table 4 Measurements overview. For aircraft campaigns, the location of the airfield is given unless no specific base can be defined (denoted by ∗ ).

Any difference in the prescribed anthropogenic and biomass burning emissions affects the atmospheric burden, the vertical layering and deposition of BC aerosol, as shown before. The corresponding uncertainties of the DRE of BC in the atmosphere and those of BC in snow are explored using the calculation method described in Sect. 2.4 . We consider the top-of-atmosphere (TOA) DRE to estimate the impact on the atmospheric radiative balance and therefore the Arctic climate. The effect at the surface (bottom of atmosphere; BOA) is considered mainly because of the implications on surface temperatures and sea-ice melting. The multi-year average TOA DRE of atmospheric BC for the BCRUS run is shown for all-sky conditions (cloudy and non-cloudy) and the years 2005–2009 in Fig. 14 a. Positive values of more than 0.2 W m −2 are calculated across the whole Arctic, indicating a net energy gain for the Arctic climate system. Values of more than 0.4 W m −2 are reached over the Arctic Ocean and the Russian Arctic. Averaged over the Arctic (60–90 ∘ N), we estimate the net DRE of atmospheric BC at 0.3 W m −2 (see Table 5 ).

Since most of the effect results from the solar spectral range, the DRE is stronger in summer and close to zero in winter. At the surface, the DRE of atmospheric BC is negative, as shown in Fig. 14 e, due to the absorption of incoming solar radiation by BC in upper atmospheric layers, which reduces the amount of energy reaching the surface. This negative effect is, however, smaller for the central Arctic Ocean than anywhere else in the Arctic, at −0.05 to −0.1 W m −2 .

Table 5 Arctic (60–90 ∘ N) field means of TOA DRE of BC averaged over the years 2005–-2009 (in W m −2 ) for the different emission setups. The terrestrial effect of in-snow BC is not calculated.

The BC-in-snow albedo effect for all-sky conditions is shown in Fig. 14 b and f, as the 2005–2009 multi-year annual mean, for TOA and surface, respectively. The difference between TOA and surface is small and mainly caused by clouds. The effect is largest in coastal Greenland at around 1 W m −2 , where snow is present throughout the year. Over the temporarily sea-ice- and snow-covered Arctic Ocean, the albedo effect varies by around 0.2 W m −2 , which compensates the negative DRE of atmospheric BC at the BOA. On average the BC-in-snow albedo effect is 0.1 W m −2 in the Arctic (60–90 ∘ N; see Table 5 ). The sum of the DRE of BC in the atmosphere and snow is shown in Fig. 14 c and g for the TOA and surface, respectively. Over the temporarily sea-ice-covered Arctic Ocean the BOA DRE of all BC (in the snow and atmosphere) is slightly positive (around 0.1 W m −2 ), while the TOA DRE is strongly positive, with values up to 1.9 W m −2 . The resulting average for the Arctic region is 0.5 W m −2 . Over the Arctic Ocean the DRE of atmospheric BC is in the range of the DRE considering all aerosol species (not shown) but smaller over the continents. The all-aerosol DRE at the TOA would therefore be negative if no BC were present in the Arctic atmosphere ( −0.2 W m −2 in the spatial and annual average).

The difference between the model runs is used to estimate the emission-related uncertainty of the Arctic energy budget. Therefore, difference of the total radiative effect at TOA (all-sky conditions) of ACCMIP-GFAS minus BCRUS, as shown in Fig. 14 h, is analyzed. In the ACCMIP-GFAS run, the TOA net all-sky positive radiative effect of BC is lower by 0.1 W m −2 in the regional average (60–90 ∘ N; see Table 5 ) but more than 0.2 W m −2 higher regionally over the Barents Sea and Kara Sea. At the surface the difference is smaller, with values of 0.05 W m −2 less in ACCMIP-GFAS over most of the Arctic, with the exception of parts of Russia, as shown in Fig. 14 h. The more recent and transient emission data with local refinement therefore result in a considerably stronger climate forcing due to anthropogenic and biomass burning BC. This shows that the TOA DRE of BC is more sensitive to an increase in the BC burden due to the different emission setups than the BOA DRE, since the net energy gain caused by the reduction of the snow albedo is canceled out to some degree by the shadowing effect of atmospheric BC.

https://www.atmos-chem-phys.net/19/11159/2019/acp-19-11159-2019-f14

Figure 14 Multi-year mean all-sky direct aerosol radiative effect (DRE) of BC for the period 2005–2009. Top row for top of the atmosphere (TOA) and bottom row for bottom of the atmosphere (BOA). (a) and (e) show the BCRUS net (terrestrial and solar) DRE of atmospheric BC, and (b) and (f) show solar BC-in-snow albedo radiative effect. (c) and (g) show the total of the radiative effects of BC in the atmosphere and deposited in snow (terrestrial plus solar). (d) and (e) show the difference in the total BC radiative effect between the runs ACCMIP-GFAS and BCRUS (ACCMIP-GFAS minus BCRUS).

We therefore conclude that, according to our best estimate, BC causes a net energy gain for the Arctic on the annual mean at TOA as well as BOA. The uncertainty with respect to the emission setup is roughly 25 % for TOA and BOA but stronger in absolute values at TOA. This is solely due to the uncertainties in emission; potential uncertainties in removal shown in the evaluation with observations are not included.

In this study, the representation of Arctic black carbon (BC) aerosol particles in the global aerosol-climate model ECHAM6.3-HAM2.3 is evaluated with respect to different emission inventories. As a reference BC measurements at Arctic sites and from aircraft campaigns are used comprehensively. By comparing the effects of different state-of-the-art BC emission inventories, an uncertainty range of current model estimates of the Arctic atmospheric BC burden and the local direct aerosol radiative effect (DRE) of BC is quantified. The uncertainties are explored with a focus on three influencing factors: (1) the influence of temporally variable biomass burning emissions, (2) the importance of recent air quality policies and economic developments, and (3) the potential improvements by regional refinements in Russian BC sources. This is achieved by comparing four different emission setups.

The run BCRUS represents a recent estimate of global emissions with the special feature of a high estimate in local Arctic emissions, especially in gas flaring. It uses anthropogenic emissions from the ECLIPSE emission data set, and in Russia the BC emissions of ECLIPSE are replaced with the higher-resolution and more recent data from Huang et al. ( 2015 ) . For the biomass burning emissions, GFAS is used, which derives the location and amount of emitted gas and aerosol particles from satellite. The ECLIPSE run uses ECLIPSE emissions and GFAS emissions for the biomass burning emissions. For the ACCMIP run we use the anthropogenic part of the ACCMIP emissions, which are widely used. We fixed the emissions to year 2000, not taking into account the recent economic changes and variable biomass burning emissions. ACCMIP does not consider gas flaring emissions. In the run ACCMIP-GFAS, the fixed year 2000 biomass burning emissions are replaced by dynamic real-time fire data from GFAS. The emission factor of 3.4 that is commonly used for GFAS emissions ( Kaiser et al. , 2012 ) was not used, since it led to a strong overestimation in mid- and high-latitudinal BC concentrations in an early setup.

The comparison between ACCMIP and ACCMIP-GFAS is used to estimate the impact of temporally variable biomass burning emissions. ACCMIP-GFAS and BCRUS are used to quantify the impact of recent developments in air quality policies and economic developments. The difference between ECLIPSE and BCRUS shows the impact of a regional refinement.

The variable biomass burning emissions are not particularly important for the annual mean of the Arctic BC burden but are crucial for reproducing high-pollution events. The different assumptions on anthropogenic emission based on economic development and air quality policies result in an uncertainty in the BC burden of more than 50 µ g m −2 over the Arctic Ocean, which is 20 % of the local annual mean BC load. The regional refinements in Russia mainly change the BC burden in this region and will improve the ability of the model to reproduce local measurements.

The near-surface BC concentrations could be reproduced to a reasonable accuracy by ECHAM-HAM in most cases. The exception from this are stations that are challenging because of their surrounding orography and the horizontal model resolution, namely Summit, Ny-Ålesund and Zeppelin Station, where ECHAM-HAM falsely produced similar peak concentrations in late winter and early spring as for all other stations. The sensitivity to the different emission setups is low in the summer. This is a result of low local emissions near the measurement sites in all runs and reduced long-range transport from the mid-latitudes as well as more precipitation in the summertime Arctic.

In the months with high modeled concentrations the model shows a high sensitivity to the changing emissions for the stations closest to the Arctic Ocean. The observed monthly median BC peak concentrations in Tiksi were underestimated by the model, but the run BCRUS that includes the most accurate gas flaring emissions produced the best results. For other stations, e.g., in Barrow in February, BCRUS showed a stronger overestimation than the other runs.

A similar pattern can be observed for Zeppelin Station, Ny-Ålesund, Villum Research Station and Alert. Higher emissions lead to higher concentrations, with no significant changes in the pattern of the annual cycle. Overall, however, it is difficult to decide which emission setup provides satisfactory agreement with the aerosol observations for all cases. This means that the annual cycle of Arctic stations reproduced by ECHAM-HAM is mainly controlled by the transport. Changing the amount and location by using a different emission setup only modulates the amount of the BC concentrations but unexpectedly does not affect the seasonality significantly.

The correlation coefficients of near-surface concentrations are generally reasonably good, at 0.45 and higher for most stations. This points toward a good agreement in the timing, especially of observed peak events. These peaks are most often caused by biomass burning. The exceptions are Summit, Simeonof, Zeppelin Station and Oulanka, with correlation coefficients below 0.3. The run ACCMIP is the only one that shows significantly smaller correlation coefficients, since the biomass burning emissions for this run are fixed and not prescribed on a daily basis from satellite observations.

The evaluation using a combination of aircraft campaigns shows that, in general, the vertical distribution is reproduced well by ECHAM-HAM. This improvement over older model versions is at least partly achieved with the aerosol size-dependent wet removal scheme by Croft et al. ( 2010 ) . The model results look best during spring. In summer BC is systematically overestimated by the model at heights above 500 hPa. This overestimation has been described for several models in the AeroCom model intercomparison project before ( Schwarz et al. , 2013 , 2017 ).

In one summer case of an observed wet removal affecting a biomass burning plume, described by Matsui et al. ( 2011 ) , the model correctly reproduced the time and height of a biomass burning layer. It is known that reproducing individual pollution events in exactly the correct way is impossible for a global model with this resolution because both the aerosol transport and the wet removal are affected by subscale processes. ECHAM-HAM overestimated the BC concentrations because of this issue. While here the BC lifetime was overestimated, in general, the BC lifetime of ECHAM-HAM was considered to be reasonably good ( Lund et al. , 2018 ) .

The ECHAM-HAM simulations show that over the Arctic Ocean the net (solar plus terrestrial) TOA DRE of atmospheric BC is positive, with an annual average of over 0.4 W m −2 . The BC-in-snow albedo effect causes an additional energy gain for the Arctic system of around 0.2 W m −2 over the central Arctic. Locally larger effects are calculated for coastal Greenland. The BOA DRE is stronger than the shadowing effect of BC, causing a net energy gain. The emission-related uncertainty of DRE both at TOA and BOA is roughly 25 %.

Overall, the current model version of ECHAM6-HAM2 performs considerably better than in a previous model intercomparison study ( Schwarz et al. , 2017 ) . In particular, the seasonality, but also the vertical distribution of BC aerosol in the Arctic, has improved. Reducing the overestimation of upper-level BC concentrations would be a big improvement, since this still causes large uncertainties in climate models and recent direct radiative forcing estimates. Here, especially the representation of wet scavenging and convective mixing needs to be improved, since it is the biggest BC sink in the Arctic.

The code for ECHAM-HAM is available to the scientific community according to the HAMMOZ Software License Agreement though the following project website: https://redmine.hammoz.ethz.ch/projects/hammoz ( Hammoz , 2019 ) . The model output data ( Schacht et al. , 2019 ) used for the plots are available through the World Data Center PANGAEA.

JS performed the ECHAM-HAM simulations, collected emission data and in situ measurement data from the providers, prepared the emissions, performed the analysis, and wrote the paper. BH provided support for the ECHAM-HAM simulations, suggested in situ measurement data providers, and provided advice during the analysis and on the project design. JQ, MZ, AE, JB and RC provided support in writing and designing the paper. RC gave advice on the emission data setup. WTKH provided the code and advice on the BC in snow parameterization for ECHAM-HAM. JB, AH, YK, AM, PRS, BW and MZ provided in situ measurement data and associated discussion. IT provided advice throughout the project design, setup, analysis and writing progress.

The authors declare that they have no conflict of interest.

We gratefully acknowledge the funding by the Deutsche Forschungsgemeinschaft (DFG; German Research Foundation; project number 268020496; TRR 172) within the Transregional Collaborative Research Centre “ArctiC Amplification: Climate Relevant Atmospheric and SurfaCe Processes, and Feedback Mechanisms (AC) 3 ”. Ribu Cherian was supported by the DFG project under grant agreement no. 637230. Bernadett Weinzierl received funding from the European Research Council under the European Community's Horizon 2020 research and innovation framework program under grant agreement no. 640458 (A-LIFE). The DLR SP2 data were obtained with the support of the European Union under grant agreement no. 265863 (ACCESS) and the Helmholtz Association under grant agreement VH-NG-606 (Helmholtz-Hochschul-Nachwuchsgruppe AerCARE). The Arctic data used in this article and managed by ACTRIS are archived and accessible from the EBAS database operated at the Norwegian Institute for Air Research (NILU; http://ebas.nilu.no , last access: 17 December 2018). ACTRIS data management is provided by the WMO Global Atmosphere Watch World Data Centre for Aerosols. The ACTRIS project, providing the data, has received funding from the European Union's Horizon 2020 research and innovation program under grant agreement no. 654109 (ACTRIS). We thank the providers of the Arctic data. ALERT data are from Environment and Climate Change Canada, Sangeeta Sharma, all technicians and operators, and the Canadian Department of National Defence. BARROW data are from Patrick Sheridan, Elisabeth Andrews and Betsy Andrews (NOAA Oceanic and Atmospheric Research/GMD). SUMMIT data are from Patrick Sheridan, Elisabeth Andrews and Betsy Andrews (NOAA Oceanic and Atmospheric Research/GMD); Michael Bergin (Duke University); and the National Science Foundation (OPP 1546002). TIKSI data are from Sara Morris (NOAA Oceanic and Atmospheric Research/GMD) and the Academy of Finland project Greenhouse gas, aerosol and albedo variations in the changing Arctic (project number 269095). PALLAS data are from the Academy of Finland project Greenhouse gas, aerosol and albedo variations in the changing Arctic (project number 269095); the Academy of Finland project Novel Assessment of Black Carbon in the Eurasian Arctic: From Historical Concentrations and Sources to Future Climate Impacts (NABCEA), project number 296302; and the Academy of Finland Centre of Excellence Programme (project number 307331). ZEPPELIN data are from the Swedish Environmental Protection Agency (Naturvårdsverket), Vetenskaprådet, FORMAS, the NILU (Norsk institutt for luftforskning) and Peter Tunved (Stockholm University). The research leading to these results has received funding from the European Union's Horizon 2020 research and innovation program under grant agreement no. 654109. IMPROVE is a collaborative association of state, tribal and federal agencies and international partners. The US Environmental Protection Agency is the primary funding source, with contracting and research support from the National Park Service. The Air Quality Group at the University of California, Davis, is the central analytical laboratory, with ion analysis provided by Research Triangle Institute and carbon analysis provided by the Desert Research Institute. We thank the principal investigators Brent Holben, Ihab Abboud, Antti Arola, Vitali Fioletov, Laurie Gregory, Rigel Kivi, Lynn Ma, Norm O’Neill, Mikhail Panchenko, Piotr Sobolewski, John R. Vande Castle and Rick Wagener; the co-investigators Piotr Glowacki, Grzegorz Karasiski and Sergey Sakerin; and their staff for establishing and maintaining the AERONET sites used in this investigation. The HIPPO 1–5 data were provided by NCAR/EOL under the sponsorship of the National Science Foundation ( https://data.eol.ncar.edu/ , last access: 12 June 2018). We would also like to thank the German Climate Computing Center (Deutsches Klimarechenzentrum; DKRZ) for the computing time and their services. We especially thank the developers of ECHAM-HAM. The ECHAM-HAMMOZ model is developed by a consortium composed of ETH Zürich, Max Planck Institute for Meteorology, Forschungszentrum Jülich, the University of Oxford, the Finnish Meteorological Institute and the Leibniz Institute for Tropospheric Research and managed by the Center for Climate Systems Modeling (C2SM) at ETH Zürich. Finally, we would like to thank the anonymous reviewers of this article for their constructive and valuable comments.

This paper was edited by Kari Lehtinen and reviewed by three anonymous referees.

Arnold, S., Law, K., Brock, C., Thomas, J., Starkweather, S., Salzen, K. V., Stohl, A., Sharma, S., Lund, M., Flanner, M., Petäjä, T., Tanimoto, H., Gamble, J., Dibb, J., Melamed, M., Johnson, N., Fidel, M., Tynkkynen, V.-P., Baklanov, A., Eckhardt, S., Monks, S., Browse, J., and Bozem, H.: Arctic air pollution: Challenges and opportunities for the next decade, ELEMENTA – Science of the Antropocene, 4, 000104, https://doi.org/10.12952/journal.elementa.000104 , 2016. a

Backman, J., Schmeisser, L., Virkkula, A., Ogren, J. A., Asmi, E., Starkweather, S., Sharma, S., Eleftheriadis, K., Uttal, T., Jefferson, A., Bergin, M., Makshtas, A., Tunved, P., and Fiebig, M.: On Aethalometer measurement uncertainties and an instrument correction factor for the Arctic, Atmos. Meas. Tech., 10, 5039–5062, https://doi.org/10.5194/amt-10-5039-2017 , 2017a. a

Backman, J., Schmeisser, L., Virkkula, A., Ogren, J. A., Asmi, E., Starkweather, S., Sharma, S., Eleftheriadis, K., Vratolis, S., Uttal, T., Tunved, P., Jefferson, A., Bergin, M., Makshtas, A., and Fiebig, M.: Time series of aerosol light-absorption coefficients from Aethalometers at six Arctic stations b etween 2012 and 2014, data retrieved from World Development Indicators, https://doi.org/10.21336/gen.1 , 2017b. a , b

Bond, T. C., Streets, D. G., Yarber, K. F., Nelson, S. M., Woo, J.-H., and Klimont, Z.: A technology-based global inventory of black and organic carbon emissions from combustion, J. Geophys. Res.-Atmos., 109, D14, https://doi.org/10.1029/2003JD003697 , 2004. a

Bond, T. C., Doherty, S. J., Fahey, D. W., Forster, P. M., Berntsen, T., DeAngelo, B. J., Flanner, M. G., Ghan, S., Kärcher, B., Koch, D., Kinne, S., Kondo, Y., Quinn, P. K., Sarofim, M. C., Schultz, M. G., Schulz, M., Venkataraman, C., Zhang, H., Zhang, S., Bellouin, N., Guttikunda, S. K., Hopke, P. K., Jacobson, M. Z., Kaiser, J. W., Klimont, Z., Lohmann, U., Schwarz, J. P., Shindell, D., Storelvmo, T., Warren, S. G., and Zender, C. S.: Bounding the role of black carbon in the climate system: A scientific assessment, J. Geophys. Res.-Atmos., 118, 5380–5552, https://doi.org/10.1002/jgrd.50171 , 2013. a

Chow, J. C., Watson, J. G., Chen, L.-W. A., Chang, M. O., Robinson, N. F., Trimble, D., and Kohl, S.: The IMPROVE_A Temperature Protocol for Thermal/Optical Carbon Analysis: Maintaining Consistency with a Long-Term Database, J. Air Waste Manage., 57, 1014–1023, https://doi.org/10.3155/1047-3289.57.9.1014 , 2007. a

Corbett, J. J., Lack, D. A., Winebrake, J. J., Harder, S., Silberman, J. A., and Gold, M.: Arctic shipping emissions inventories and future scenarios, Atmos. Chem. Phys., 10, 9689–9704, https://doi.org/10.5194/acp-10-9689-2010 , 2010. a

Croft, B., Lohmann, U., Martin, R. V., Stier, P., Wurzler, S., Feichter, J., Hoose, C., Heikkilä, U., van Donkelaar, A., and Ferrachat, S.: Influences of in-cloud aerosol scavenging parameterizations on aerosol concentrations and wet deposition in ECHAM5-HAM, Atmos. Chem. Phys., 10, 1511–1543, https://doi.org/10.5194/acp-10-1511-2010 , 2010. a

Eckhardt, S., Quennehen, B., Olivié, D. J. L., Berntsen, T. K., Cherian, R., Christensen, J. H., Collins, W., Crepinsek, S., Daskalakis, N., Flanner, M., Herber, A., Heyes, C., Hodnebrog, Ø., Huang, L., Kanakidou, M., Klimont, Z., Langner, J., Law, K. S., Lund, M. T., Mahmood, R., Massling, A., Myriokefalitakis, S., Nielsen, I. E., Nøjgaard, J. K., Quaas, J., Quinn, P. K., Raut, J.-C., Rumbold, S. T., Schulz, M., Sharma, S., Skeie, R. B., Skov, H., Uttal, T., von Salzen, K., and Stohl, A.: Current model capabilities for simulating black carbon and sulfate concentrations in the Arctic atmosphere: a multi-model evaluation using a comprehensive measurement data set, Atmos. Chem. Phys., 15, 9413–9433, https://doi.org/10.5194/acp-15-9413-2015 , 2015. a , b , c

Eckhardt, S., Cassiani, M., Evangeliou, N., Sollum, E., Pisso, I., and Stohl, A.: Source-receptor matrix calculation for deposited mass with the Lagrangian particle dispersion model FLEXPART v10.2 in backward mode, Geosci. Model Dev., 10, 4605–4618, https://doi.org/10.5194/gmd-10-4605-2017 , 2017. a

Flanner, M. G., Zender, C. S., Randerson, J. T., and Rasch, P. J.: Present-day climate forcing and response from black carbon in snow, J. Geophys. Res.-Atmos., 112, D11202, https://doi.org/10.1029/2006JD008003 , 2007. a , b

Gilgen, A., Huang, W. T. K., Ickes, L., Neubauer, D., and Lohmann, U.: How important are future marine and shipping aerosol emissions in a warming Arctic summer and autumn?, Atmos. Chem. Phys., 18, 10521–10555, https://doi.org/10.5194/acp-18-10521-2018 , 2018. a , b , c

Hammoz: ECHAM-HAMMOZ, available at: https://redmine.hammoz.ethz.ch/projects/hammoz , last access: 9 May 2019. a

Herber, A. B., Haas, C., Stone, R. S., Bottenheim, J. W., Liu, P., Li, S.-M., Staebler, R. M., Strapp, J. W., and Dethloff, K.: Regular airborne surveys of Arctic sea ice and atmosphere, Eos, Transactions American Geophysical Union, 93, 41–42, https://doi.org/10.1029/2012EO040001 , 2012. a , b

Huang, K., Fu, J. S., Prikhodko, V. Y., Storey, J. M., Romanov, A., Hodson, E. L., Cresko, J., Morozova, I., Ignatieva, Y., and Cabaniss, J.: Russian anthropogenic black carbon: Emission reconstruction and Arctic black carbon simulation, J. Geophys. Res.-Atmos., 120, 11306–11333, https://doi.org/10.1002/2015JD023358 , 2015. a , b , c , d , e , f

Ikeda, K., Tanimoto, H., Sugita, T., Akiyoshi, H., Kanaya, Y., Zhu, C., and Taketani, F.: Tagged tracer simulations of black carbon in the Arctic: transport, source contributions, and budget, Atmos. Chem. Phys., 17, 10515–10533, https://doi.org/10.5194/acp-17-10515-2017 , 2017. a

Jacob, D. J., Crawford, J. H., Maring, H., Clarke, A. D., Dibb, J. E., Emmons, L. K., Ferrare, R. A., Hostetler, C. A., Russell, P. B., Singh, H. B., Thompson, A. M., Shaw, G. E., McCauley, E., Pederson, J. R., and Fisher, J. A.: The Arctic Research of the Composition of the Troposphere from Aircraft and Satellites (ARCTAS) mission: design, execution, and first results, Atmos. Chem. Phys., 10, 5191–5212, https://doi.org/10.5194/acp-10-5191-2010 , 2010. a , b

Kaiser, J. W., Heil, A., Andreae, M. O., Benedetti, A., Chubarova, N., Jones, L., Morcrette, J.-J., Razinger, M., Schultz, M. G., Suttie, M., and van der Werf, G. R.: Biomass burning emissions estimated with a global fire assimilation system based on observed fire radiative power, Biogeosciences, 9, 527–554, https://doi.org/10.5194/bg-9-527-2012 , 2012. a , b , c

Klimont, Z., Kupiainen, K., Heyes, C., Purohit, P., Cofala, J., Rafaj, P., Borken-Kleefeld, J., and Schöpp, W.: Global anthropogenic emissions of particulate matter including black carbon, Atmos. Chem. Phys., 17, 8681–8723, https://doi.org/10.5194/acp-17-8681-2017 , 2017. a

Koch, D., Schulz, M., Kinne, S., McNaughton, C., Spackman, J. R., Balkanski, Y., Bauer, S., Berntsen, T., Bond, T. C., Boucher, O., Chin, M., Clarke, A., De Luca, N., Dentener, F., Diehl, T., Dubovik, O., Easter, R., Fahey, D. W., Feichter, J., Fillmore, D., Freitag, S., Ghan, S., Ginoux, P., Gong, S., Horowitz, L., Iversen, T., Kirkevåg, A., Klimont, Z., Kondo, Y., Krol, M., Liu, X., Miller, R., Montanaro, V., Moteki, N., Myhre, G., Penner, J. E., Perlwitz, J., Pitari, G., Reddy, S., Sahu, L., Sakamoto, H., Schuster, G., Schwarz, J. P., Seland, Ø., Stier, P., Takegawa, N., Takemura, T., Textor, C., van Aardenne, J. A., and Zhao, Y.: Evaluation of black carbon estimations in global aerosol models, Atmos. Chem. Phys., 9, 9001–9026, https://doi.org/10.5194/acp-9-9001-2009 , 2009. a

Kodros, J. K., Hanna, S. J., Bertram, A. K., Leaitch, W. R., Schulz, H., Herber, A. B., Zanatta, M., Burkart, J., Willis, M. D., Abbatt, J. P. D., and Pierce, J. R.: Size-resolved mixing state of black carbon in the Canadian high Arctic and implications for simulated direct radiative effect, Atmos. Chem. Phys., 18, 11345–11361, https://doi.org/10.5194/acp-18-11345-2018 , 2018. a

Lamarque, J.-F., Bond, T. C., Eyring, V., Granier, C., Heil, A., Klimont, Z., Lee, D., Liousse, C., Mieville, A., Owen, B., Schultz, M. G., Shindell, D., Smith, S. J., Stehfest, E. , Van Aardenne, J., Cooper, O. R., Kainuma, M., Mahowald, N., McConnell, J. R., Naik, V., Riahi, K., and van Vuuren, D. P.: Historical (1850–2000) gridded anthropogenic and biomass burning emissions of reactive gases and aerosols: methodology and application, Atmos. Chem. Phys., 10, 7017–7039, https://doi.org/10.5194/acp-10-7017-2010 , 2010. a

Law, K. S. and Stohl, A.: Arctic Air Pollution: Origins and Impacts, Science, 315, 1537–1540, https://doi.org/10.1126/science.1137695 , 2007. a , b

Lohmann, U., Spichtinger, P., Jess, S., Peter, T., and Smit, H. G. J.: Cirrus cloud formation and ice supersaturated regions in a global climate model, Environ. Res. Lett., 3, 1748–9326, https://doi.org/10.1088/1748-9326/3/4/045022 , 2008. a

Lund, M. T., Samset, B. r. H., Skeie, R. B., Watson-Parris, D., Katich, J. M., Schwarz, J. P., and Weinzierl, B.: Short Black Carbon lifetime inferred from a global set of aircraft observations, npj Climate and Atmospheric Science, 1, 2397–3722 https://doi.org/10.1038/s41612-018-0040-x , 2018. a , b , c

Massling, A., Nielsen, I. E., Kristensen, D., Christensen, J. H., Sørensen, L. L., Jensen, B., Nguyen, Q. T., Nøjgaard, J. K., Glasius, M., and Skov, H.: Atmospheric black carbon and sulfate concentrations in Northeast Greenland, Atmos. Chem. Phys., 15, 9681–9692, https://doi.org/10.5194/acp-15-9681-2015 , 2015. a , b

Matsui, H., Kondo, Y., Moteki, N., Takegawa, N., Sahu, L. K., Koike, M., Zhao, Y., Fuelberg, H. E., Sessions, W. R., Diskin, G., Anderson, B. E., Blake, D. R., Wisthaler, A., Cubison, M. J., and Jimenez, J. L.: Accumulation-mode aerosol number concentrations in the Arctic during the ARCTAS aircraft campaign: Long-range transport of polluted and clean air from the Asian continent, J. Geophys. Res.-Atmos., 116, D20, https://doi.org/10.1029/2011JD016189 , 2011. a , b , c , d

Petoukhov, V. and Semenov, V. A.: A link between reduced Barents-Kara sea ice and cold winter extremes over northern continents, J. Geophys. Res.-Atmos., 115, D21111, https://doi.org/10.1029/2009JD013568 , 2010. a

Quinn, P., AMAP, and Stohl, A.: AMAP Assessment 2015: Black Carbon and Ozone as Arctic Climate Forcers, Arctic Monitoring and Assessment Programme (AMAP), Oslo, Norway, available at: https://www.amap.no/documents/download/2506 (last access: 5 May 2017), 2015. a

Roiger, A., Thomas, J.-L., Schlager, H., Law, K. S., Kim, J., Schäfler, A., Weinzierl, B., Dahlkötter, F., Krisch, I., Marelle, L., Minikin, A., Raut, J.-C., Reiter, A., Rose, M., Scheibe, M., Stock, P., Baumann, R., Bouarar, I., Clerbaux, C., George, M., Onishi, T., and Flemming, J.: Quantifying Emerging Local Anthropogenic Emissions in the Arctic Region: The ACCESS Aircraft Campaign Experiment, B. Am. Meteorol. Soc., 96, 441–460, https://doi.org/10.1175/BAMS-D-13-00169.1 , 2015. a , b

Samset, B. H., Myhre, G., Schulz, M., Balkanski, Y., Bauer, S., Berntsen, T. K., Bian, H., Bellouin, N., Diehl, T., Easter, R. C., Ghan, S. J., Iversen, T., Kinne, S., Kirkevåg, A., Lamarque, J.-F., Lin, G., Liu, X., Penner, J. E., Seland, Ø., Skeie, R. B., Stier, P., Takemura, T., Tsigaridis, K., and Zhang, K.: Black carbon vertical profiles strongly affect its radiative forcing uncertainty, Atmos. Chem. Phys., 13, 2423–2434, https://doi.org/10.5194/acp-13-2423-2013 , 2013. a

Sand, M., Berntsen, T. K., Kay, J. E., Lamarque, J. F., Seland, Ø., and Kirkevåg, A.: The Arctic response to remote and local forcing of black carbon, Atmos. Chem. Phys., 13, 211–224, https://doi.org/10.5194/acp-13-211-2013 , 2013. a

Sand, M., Berntsen, T., von Salzen, K., Flanner, M., Langner, J., and G. Victor, D.: Response of Arctic temperature to changes in emissions of short-lived climate forcers, 6, 286–289, https://doi.org/10.1038/nclimate2880, 2015. a

Sand, M., Samset, B. H., Balkanski, Y., Bauer, S., Bellouin, N., Berntsen, T. K., Bian, H., Chin, M., Diehl, T., Easter, R., Ghan, S. J., Iversen, T., Kirkevåg, A., Lamarque, J.-F., Lin, G., Liu, X., Luo, G., Myhre, G., Noije, T. V., Penner, J. E., Schulz, M., Seland, Ø., Skeie, R. B., Stier, P., Takemura, T., Tsigaridis, K., Yu, F., Zhang, K., and Zhang, H.: Aerosols at the poles: an AeroCom Phase II multi-model evaluation, Atmos. Chem. Phys., 17, 12197–12218, https://doi.org/10.5194/acp-17-12197-2017 , 2017. a , b

Schacht, J., Heinold, B., and Tegen, I.: ECHAM-HAM simulations for evaluation and quantification of BC effects in the Arctic under different emission data sets, Leibniz-Institut für Troposphärenforschung e.V., Leipzig, PANGAEA, available at: https://doi.org/10.1594/PANGAEA.903547 , last access: 6 July 2019. a

Schutgens, N. A. J., Gryspeerdt, E., Weigum, N., Tsyro, S., Goto, D., Schulz, M., and Stier, P.: Will a perfect model agree with perfect observations? The impact of spatial sampling, Atmos. Chem. Phys., 16, 6335–6353, https://doi.org/10.5194/acp-16-6335-2016 , 2016. a

Schwarz, J. P., Samset, B. H., Perring, A. E., Spackman, J. R., Gao, R. S., Stier, P., Schulz, M., Moore, F. L., Ray, E. A., and Fahey, D. W.: Global-scale seasonally resolved black carbon vertical profiles over the Pacific, Geophys. Res. Lett., 40, 5542–5547, https://doi.org/10.1002/2013GL057775 , 2013. a , b , c

Schwarz, J. P., Weinzierl, B., Samset, B. r. H., Dollner, M., Heimerl, K., Markovic, M. Z., Perring, A. E., and Ziemba, L.: Aircraft measurements of black carbon vertical profiles show upper tropospheric variability and stability, Geophys. Res. Lett., 44, 1132–1140, https://doi.org/10.1002/2016GL071241 , 2017. a , b

Shaw, G. E.: The Arctic Haze Phenomenon, B. Am. Meteorol. Soc., 76, 2403–2414, https://doi.org/10.1175/1520-0477(1995)076<2403:TAHP>2.0.CO;2 , 1995. a

Shindell, D.: Local and remote contributions to Arctic warming, Geophys. Res. Lett., 34, L14704, https://doi.org/10.1029/2007GL030221 , 2007. a

Shindell, D. T., Chin, M., Dentener, F., Doherty, R. M., Faluvegi, G., Fiore, A. M., Hess, P., Koch, D. M., MacKenzie, I. A., Sanderson, M. G., Schultz, M. G., Schulz, M., Stevenson, D. S., Teich, H., Textor, C., Wild, O., Bergmann, D. J., Bey, I., Bian, H., Cuvelier, C., Duncan, B. N., Folberth, G., Horowitz, L. W., Jonson, J., Kaminski, J. W., Marmer, E., Park, R., Pringle, K. J., Schroeder, S., Szopa, S., Takemura, T., Zeng, G., Keating, T. J., and Zuber, A.: A multi-model assessment of pollution transport to the Arctic, Atmos. Chem. Phys., 8, 5353–5372, https://doi.org/10.5194/acp-8-5353-2008 , 2008. a

Shindell, D. T., Faluvegi, G., Koch, D. M., Schmidt, G. A., Unger, N., and Bauer, S. E.: Improved Attribution of Climate Forcing to Emissions, Science, 326, 716–718, https://doi.org/10.1126/science.1174760 , 2009. a

Sinha, P. R., Kondo, Y., Koike, M., Ogren, J. A., Jefferson, A., Barrett, T. E., Sheesley, R. J., Ohata, S., Moteki, N., Coe, H., Liu, D., Irwin, M., Tunved, P., Quinn, P. K., and Zhao, Y.: Evaluation of ground-based black carbon measurements by filter- based photometers at two Arctic sites, J. Geophys. Res.-Atmos., 122, 3544–3572, https://doi.org/10.1002/2016JD025843 , 2017. a , b

Sofiev, M., Vankevich, R., Lotjonen, M., Prank, M., Petukhov, V., Ermakova, T., Koskinen, J., and Kukkonen, J.: An operational system for the assimilation of the satellite information on wild-land fires for the needs of air quality modelling and forecasting, Atmos. Chem. Phys., 9, 6833–6847, https://doi.org/10.5194/acp-9-6833-2009 , 2009. a

Stevens, B., Giorgetta, M., Esch, M., Mauritsen, T., Crueger, T., Rast, S., Salzmann, M., Schmidt, H., Bader, J., Block, K., Brokopf, R., Fast, I., Kinne, S., Kornblueh, L., Lohmann, U., Pincus, R., Reichler, T., and Roeckner, E.: Atmospheric component of the MPI-M Earth System Model: ECHAM6, J. Adv. Model. Earth Sy., 5, 146–172, https://doi.org/10.1002/jame.20015 , 2013. a

Stier, P., Feichter, J., Kinne, S., Kloster, S., Vignati, E., Wilson, J., Ganzeveld, L., Tegen, I., Werner, M., Balkanski, Y., Schulz, M., Boucher, O., Minikin, A., and Petzold, A.: The aerosol-climate model ECHAM5-HAM, Atmos. Chem. Phys., 5, 1125–1156, https://doi.org/10.5194/acp-5-1125-2005 , 2005. a , b , c

Stohl, A., Klimont, Z., Eckhardt, S., Kupiainen, K., Shevchenko, V. P., Kopeikin, V. M., and Novigatsky, A. N.: Black carbon in the Arctic: the underestimated role of gas flaring and residential combustion emissions, Atmos. Chem. Phys., 13, 8833–8855, https://doi.org/10.5194/acp-13-8833-2013 , 2013. a , b , c , d , e , f , g

Tegen, I. and Heinold, B.: Large-Scale Modeling of Absorbing Aerosols and Their Semi-Direct Effects, Atmosphere, 9, 10, https://doi.org/10.3390/atmos9100380 , 2018. a

Tegen, I., Neubauer, D., Ferrachat, S., Siegenthaler-Le Drian, C., Bey, I., Schutgens, N., Stier, P., Watson-Parris, D., Stanelle, T., Schmidt, H., Rast, S., Kokkola, H., Schultz, M., Schroeder, S., Daskalakis, N., Barthel, S., Heinold, B., and Lohmann, U.: The global aerosol–climate model ECHAM6.3–HAM2.3 – Part 1: Aerosol evaluation, Geosci. Model Dev., 12, 1643–1677, https://doi.org/10.5194/gmd-12-1643-2019 , 2019. a , b

Trenberth, K. E., Jones, P. D., Ambenje, P., Bojariu, R., Easterling, D., Klein Tank, A., Parker, D., Rahimzadeh, F., Renwick, J. A., Rusticucci, M., Soden, B., and Zhai, P.: Observations: Surface and Atmospheric Climate Change, in: Climate Change 2007: The Physical Science Basis, Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, edited by: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K. B., Tignor, M., and Miller, H. L., Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2007. a

van Vuuren, D. P., Edmonds, J., Kainuma, M., Riahi, K., Thomson, A., Hibbard, K., Hurtt, G. C., Kram, T., Krey, V., Lamarque, J.-F., Masui, T., Meinshausen, M., Nakicenovic, N., Smith, S. J., and Rose, S. K.: The representative concentration pathways: an overview, Clim. Change, 109, 5, https://doi.org/10.1007/s10584-011-0148-z , 2011. a

Vignati, E., Wilson, J., and Stier, P.: M7: An efficient size-resolved aerosol microphysics module for large-scale aerosol transport models, J. Geophys. Res.-Atmos., 109, D22, https://doi.org/10.1029/2003JD004485 , 2004. a

Vignati, E., Karl, M., Krol, M., Wilson, J., Stier, P., and Cavalli, F.: Sources of uncertainties in modelling black carbon at the global scale, Atmos. Chem. Phys., 10, 2595–2611, https://doi.org/10.5194/acp-10-2595-2010 , 2010. a , b

Wendisch, M., Brückner, M., Burrows, J. P., Crewell, S., Dethloff, K., Ebell, K., Lüpkes, C., Macke, A., Notholt, J., Quaas, J., Rinke, A., and Tegen, I.: Understanding causes and effects of rapid warming in the Arctic, Eos, 98, 22–26, https://doi.org/10.1029/2017EO064803 , 2017. a

Wendisch, M., Macke, A., Ehrlich, A., Lüpkes, C., Mech, M., Chechin, D., Dethloff, K., Barientos, C., Bozem, H., Brückner, M., Clemen, H.-C., Crewell, S., Donth, T., Dupuy, R., Ebell, K., Egerer, U., Engelmann, R., Engler, C., Eppers, O., Gehrmann, M., Gong, X., Gottschalk, M., Gourbeyre, C., Griesche, H., Hartmann, J., Hartmann, M., Heinold, B., Herber, A., Herrmann, H., Heygster, G., Hoor, P., Jafariserajehlou, S., Jäkel, E., Järvinen, E., Jourdan, O., Kästner, U., Kecorius, S., Knudsen, E. M., Köllner, F., Kretzschmar, J., Lelli, L., Leroy, D., Maturilli, M., Mei, L., Mertes, S., Mioche, G., Neuber, R., Nicolaus, M., Nomokonova, T., Notholt, J., Palm, M., van Pinxteren, M., Quaas, J., Richter, P., Ruiz-Donoso, E., Schäfer, M., Schmieder, K., Schnaiter, M., Schneider, J., Schwarzenböck, A., Seifert, P., Shupe, M. D., Siebert, H., Spreen, G., Stapf, J., Stratmann, F., Vogl, T., Welti, A., Wex, H., Wiedensohler, A., Zanatta, M., and Zeppenfeld, S.: The Arctic Cloud Puzzle: Using ACLOUD/PASCAL Multi-Platform Observations to Unravel the Role of Clouds and Aerosol Particles in Arctic Amplification, B. Am. Meteorol. Soc., 100, 841–871, https://doi.org/10.1175/BAMS-D-18-0072.1 , 2018. a , b

Winiger, P., Andersson, A., Eckhardt, S., Stohl, A., Semiletov, I. P., Dudarev, O. V., Charkin, A., Shakhova, N., Klimont, Z., Heyes, C., and Gustafsson, Ö.: Siberian Arctic black carbon sources constrained by model and observation, P. Natl. Acad. Sci. USA, 7, 1054–1061, https://doi.org/10.1073/pnas.1613401114 , 2017. a , b

Wofsy, S. C., Daube, B., Jimenez, R., Kort, E., Pittman, J. V., Park, S., Commane, R., Xian g, B., Santoni, G., Jacob, D. J., Fisher, J. A., Heaps, C. A. Pickett-and Wang, H., Wecht, K. J., Wang, Q., Stephens, B. B., Shertz, S., Watt, A., Romashkin, P., Campos, T., Haggerty, J., Coo per, W. A., Rogers, D. C., Beaton, S., Hendershot, R., Elkins, J. W., Fahey, D. W., Gao, R.-S., Schwarz, J. P., Moore, F., Montzka, S. A., Perring, A. E., Hurst, D., Miller, B. R., Sweeney, C., Oltmans, S. J., Hintsa, E., Nance, D., Dutton, G. S., Watts, L. A., Spackman, J. R., Rosenlof, K. H., Ray, E., Hall, B., Zondlo, M., Diao, M., Keeling, R. F., Bent, J., Atlas, E., Lueb, R., and Mahoney, M. J.: HIPPO Merged 10-Second Meteorology, Atmospheric Chemistry, and Aerosol Dat, Version 1.0., dat retrieved from UCAR/NCAR – Earth Observing Laboratory, available at: https://doi.org/10.3334/CDIAC/HIPPO_010 (last access: 12 June 2018), 2017. a , b

Zanatta, M., Gysel, M., Bukowiecki, N., Müller, T., Weingartner, E., Areskoug, H., Fiebig, M., Yttri, K., Mihalopoulos, N., Kouvarakis, G., Beddows, D., Harrison, R., Cavalli, F., Putaud, J., Spindler, G., Wiedensohler, A., Alastuey, A., Pandolfi, M., Sellegri, K., Swietlicki, E., Jaffrezo, J., Baltensperger, U., and Laj, P.: A European aerosol phenomenology-5: Climatology of black carbon optical properties at 9 regional background sites across Europe, Atmos. Environ., 145, 346–364, https://doi.org/10.1016/j.atmosenv.2016.09.035 , 2016. a

Zanatta, M., Laj, P., Gysel, M., Baltensperger, U., Vratolis, S., Eleftheriadis, K., Kondo, Y., Dubuisson, P., Winiarek, V., Kazadzis, S., Tunved, P., and Jacobi, H.-W.: Effects of mixing state on optical and radiative properties of black carbon in the European Arctic, Atmos. Chem. Phys., 18, 14037–14057, https://doi.org/10.5194/acp-18-14037-2018 , 2018. a , b

Zhang, K., O'Donnell, D., Kazil, J., Stier, P., Kinne, S., Lohmann, U., Ferrachat, S., Croft, B., Quaas, J., Wan, H., Rast, S., and Feichter, J.: The global aerosol-climate model ECHAM-HAM, version 2: sensitivity to improvements in process representations, Atmos. Chem. Phys., 12, 8911–8949, https://doi.org/10.5194/acp-12-8911-2012 , 2012. a , b , c , d , e , f , g

Introduction
Methodology
Sensitivity study on emissions
Evaluation with observations
Direct aerosol radiative effects of BC
Summary and conclusions
Code and data availability
Author contributions
Competing interests
Acknowledgements
Review statement

Visual Analytics of Air Pollutant Propagation Path and Pollution Source

New citation alert added.

This alert has been successfully added and will be sent to:

You will be notified whenever a record that you have chosen has been cited.

To manage your alert preferences, click on the button below.

New Citation Alert!

Please log in to your account

Information & Contributors

Bibliometrics & citations.

Yang L Li R Qiu X Bi C (2024) BLCM: a BP-LGBM-based atmospheric visibility forecasting model Journal of Visualization 10.1007/s12650-024-01009-6 Online publication date: 5-Jun-2024 https://doi.org/10.1007/s12650-024-01009-6
Jiao C Bi C Yang L (2024) FFEINR: flow feature-enhanced implicit neural representation for spatiotemporal super-resolution Journal of Visualization 10.1007/s12650-024-00959-1 27 :2 (273-289) Online publication date: 19-Feb-2024 https://dl.acm.org/doi/10.1007/s12650-024-00959-1

Index Terms

Human-centered computing

Visualization

Visualization application domains

Visual analytics

Information systems

Information systems applications

Spatial-temporal systems

Geographic information systems

Recommendations

An analysis of the main pollution source of sewage river in north china.

Taking the main pollution source of sewage river in north China as study object, concentrations of ammonia, chemical oxygen demand and heavy metals in sources of pollution and nearby water are determined. The two dimension steady river water quality ...

Tracing-U-Net: An Attention Based U-Net model for Air Pollution Source Tracing from Sparse Dataset

In recent years, the increasingly severe air pollution has caused significant harm to us, so it is imperative to effectively manage air pollution from its source. Existing methods on pollutant tracing primarily falls into two types: numerical solution ...

Spatial-temporal Characteristics and Source Identification of Water Pollutants in Xi'an Moat

To reveal the water quality of Xi'an moat, a water quality index is developed for water quality assessment. This study collected the water quality data in Xi'an moat from 2018 to 2020. The results based on single-factor water quality assessment show ...

Information

Published in.

Association for Computing Machinery

New York, NY, United States

Publication History

Permissions, check for updates, author tags.

Pollution source
propagation path
visual analytics
Research-article
Refereed limited

Funding Sources

National Natural Science Foundation of China
the National Key R&D Program of China

Acceptance Rates

Contributors, other metrics, bibliometrics, article metrics.

2 Total Citations View Citations
43 Total Downloads
Downloads (Last 12 months) 43
Downloads (Last 6 weeks) 2

View Options

Check if you have access through your login credentials or your institution to get full access on this article.

Full Access

View options.

View or Download as a PDF file.

View online with eReader .

HTML Format

View this article in HTML Format.

Share this Publication link

Copying failed.

Share on social media

Affiliations, export citations.

Please download or close your previous search result export first before starting a new bulk export. Preview is not available. By clicking download, a status dialog will open to start the export process. The process may take a few minutes but once it finishes a file will be downloadable from your browser. You may continue to browse the DL while the export process is in progress. Download
Download citation
Copy citation

We are preparing your search results for download ...

We will inform you here when the file is ready.

Your file of search results citations is now ready.

Your search export query has expired. Please try again.

Investigating the association between air pollutants’ concentration and meteorological parameters in a rapidly growing urban center of West Bengal, India: a statistical modeling-based approach

Original Article
Published: 05 January 2023
Volume 9 , pages 2877–2892, ( 2023 )

Cite this article

Arghadeep Bose ORCID: orcid.org/0000-0002-3284-4633 1 &
Indrajit Roy Chowdhury ORCID: orcid.org/0000-0001-9491-2999 1

3365 Accesses

14 Citations

Explore all metrics

The ambient air quality in a city is heavily influenced by meteorological conditions. The city of Siliguri, known as the “Gateway of Northeast India”, is a major hotspot of air pollution in the Indian state of West Bengal. Yet almost no research has been done on the possible impacts of meteorological factors on criterion air pollutants in this rapidly growing urban area. From March 2018 to September 2022, the present study aimed to determine the correlations between meteorological factors, including daily mean temperature (℃), relative humidity (%), rainfall (mm), wind speed (m/s) with the concentration of criterion air pollutants (PM 2.5 , PM 10 , NO 2 , SO 2 , CO, O 3 , and NH 3 ). For this research, the trend of all air pollutants over time was also investigated. The Spearman correlation approach was used to correlate the concentration of air pollutants with the effect of meteorological variables on these pollutants. Comparing the multiple linear regression (MLR) and non-linear regression (MLNR) models permitted to examine the potential influence of meteorological factors on concentrations of air pollutants. According to the trend analysis, the concentration of NH3 in the air of Siliguri is rising, while the concentration of other pollutants is declining. Most pollutants showed a negative correlation with meteorological variables; however, the seasons impacted on how they responded. The comparative regression research results showed that although the linear and non-linear models performed well in predicting particulate matter concentrations, they performed poorly in predicting gaseous contaminants. When considering seasonal fluctuations and meteorological parameters, the results of this research will definitely help to increase the accuracy of air pollution forecasting near future.

Examining the relationship between atmospheric pollutants and meteorological factors in Asansol city, West Bengal, India, using statistical modelling

Air quality pollutants and their relationship with meteorological variables in four suburbs of Greater Sydney, Australia

Quantifying effects of meteorological parameters on air pollution in Kathmandu valley through regression models

Explore related subjects.

Environmental Chemistry

Avoid common mistakes on your manuscript.

Introduction

Urban air pollution is a severe and escalating environmental issue, particularly in developing nations. It also has a significant impact on global public health, notably on respiratory and cardiovascular diseases (Mage et al. 1996 ; Bernard et al. 2001 ; Shi et al. 2020 ). Despite the significant progress achieved in the prevention and management of air pollution over the past several years, the majority of inhabitants continue to be quite concerned about such crucial ambient exposure (WHO 2021 ; Leung et al. 2020 ; Molina 2021 ). Many nations in the developing world have air pollution levels that are dangerously close to, or even over, the threshold recommended by the World Health Organization (Tiwari et al. 2014 ; Doreswamy et al. 2020 ; Srivastava 2022 ). Since the industrial revolution and widespread urbanization, air pollution has risen to the top of the environmental concerns list in both developed and developing nations (Anwar et al. 2021 ; Wei et al. 2021 ; Zhang et al. 2022 ). The key source of pollutants that contribute to the degradation of air quality are various human activities, such as fossil fuel combustion to drive production processes, motor vehicles, and industrial plants (Pachón et al. 2018 ; Rajput et al. 2021 ; Munsif et al. 2021 ; Molina 2021 ). In addition, the primary factors contributing to the degradation of air quality in developing nations are the tremendous expansion of the urban population and the changes in land use carried on by urban development (Liang et al. 2019 ; Surya et al. 2020 ). Because of this, the quality of the air, both indoors and outdoors, varies from what is considered normal in urban settings. This implies that a large number of urban inhabitants are continually exposed to an unhealthy amount of air pollution (Chen et al. 2020 ). In metropolitan areas, more than 80% of people are exposed to pollutants at levels exceeding WHO standards, and 98% of metropolitan areas in middle as well as low-income countries fail to fulfill air quality standards (Manju et al. 2018 ; Glazener and Khreis 2019 ; Afghan et al. 2022 ). In 2012, there were almost 7 million air pollution-related deaths, with 4.2 million of such mortalities being directly linked to exposure to outdoor air pollution (Sharma et al. 2013 ; Kayes et al. 2019 ; Piracha and Chaudhary 2022 ).

India is experiencing problems pertaining to a growing population and poor air quality. India ranked third worldwide for PM pollution-related deaths in 2017 (Mitoma et al. 2021 ). There were about 1.1 million avoidable fatalities in India in 2017 due to air pollution (Jat and Gurjar 2021 ), with 56% attributable to outdoor PM2.5 and 44% to indoor air pollution (Pal et al. 2022 ). According to WHO ( 2016 ), 10 of the 20 most populous cities are in India. After analyzing PM2.5 emissions from all countries, WHO ( 2019 ) categorize India as the fifth most polluting country, with 21 of the top 30 most polluted cities being situated in India. Considering 2016 statistics, at least 140 million Indians breathe air ten times or more above the WHO acceptable limit (Chatterji 2021 ), while 13 of the world's top 20 cities with the greatest yearly air pollution are already in India (Agarwal et al. 2020 ; Roy and Singha 2021 ). Due to an increase in industry, population concentration, anthropogenic influences, and automobile usage, India's air quality has been worse over time (Gurjar et al. 2008 ; Jain et al. 2018 ; Jat and Gurjar 2021 ). In recent decades, greenhouse gas emissions (GHGs) and other pollutants have increased in both megacities and small to medium-sized urban centers (Kulkarni et al. 2018 ; Kumar and Gurjar 2019 ; Wen et al. 2020 ). Siliguri isn't an exception in this case (Roy and Singha 2020 ; Biswas et al. 2020 ; Halder and Bandyopadhyay ( 2022 ). Constant population growth has resulted in excessive energy consumption, which has a negative influence on the environment and the air quality in big as well as medium-sized cities like Siliguri (Bose and Chowdhury 2020 ; Roy and Singha 2020 ; Biswas et al. 2020 ).

Urban air quality is affected by meteorological factors (Gualtieri et al. 2015 ; Manju et al. 2018 ; Seo et al. 2018 ; Suhaimi et al. 2020 ; Peng et al. 2020 ; Haddad and Vizakos 2020 ). Atmospheric pollutants are affected by a variety of factors, including temperature, humidity, wind speed, and direction, including the processes of generation and diffusion (Whiteman et al. 2014 ; Hoek et al. 2015 ; Seo et al. 2018 ; Jain et al. 2018 ). They are very important for controlling pollution levels (Peng et al. 2020 ; Haddad and Vizakos 2020 ). The chemical interactions in the atmosphere between precipitation and air pollutants may remove gaseous pollutants and deposit particles, although rain's impact on air quality varies greatly (Dayan and Levy 2005 ; Elperin et al. 2011 ; Ouyang et al. 2015 ; Kayes et al. 2019 ). Evidence from numerous research indicate that air quality may indeed be altered by meteorological conditions (Jhun et al. 2015 ; Manju et al. 2018 ; Borge et al. 2019 ; Castelhano et al. 2022 ). West Bengal has seasonal changes in temperature, precipitation, and humidity due to its subtropical monsoon climate (Bhunia et al. 2019 ; Kundu 2020 ). Seasonal changes in air quality may be seen across the state (Biswas et al. 2020 ). PM2.5 and PM10 levels are higher than acceptable during the dry season but are lower during the monsoon, according to national regulations (Jain et al. 2020 ; Roy and Singha 2020 ). Numerous studies on the effects of weather have been conducted in China (Kan et al. 2012 ; Li et al. 2014 , 2020 ; He et al. 2017 ; Song et al. 2017 ; Lin et al. 2020 ; Liu and Wang 2020 ; Gao et al. 2021 ) and India especially at the national level and large cities (Ramanathan and Feng 2009 ; Bhaskar and Mehta 2010 ; Jayamurugan et al. 2013 ), but it is understudied in medium-sized urban areas like Siliguri. This research aimed to examine the links between air pollutant concentration and meteorological factors, assess seasonal fluctuations in air pollutant concentration and perform a time series assessment of air pollutant concentration in Siliguri city.

Siliguri is a rapidly expanding city with a solid foundation of trade as well as commerce and substantial economic activity. At an elevation of approximately 122 m (400 feet), it is situated in the Himalayan foothills on the banks of the river Mahananda. It has had substantial population growth in recent years, and the Siliguri Municipal Corporation's geographic area is about 41.9 km2 and is bounded by 47 wards that have grown by five times since 1931. Siliguri, in West Bengal, is indeed the third-largest urban agglomeration after Kolkata and Asansol. By luring a sizable number of migrants over the course of time, Siliguri has transformed from a small village into a financially advanced city. This amazing expansion of the city is a result of the city's tremendous population expansion (Bose and Chowdhury 2020 ). Evidence from the results of the 2011 Census showed that the city's population increased at a quicker pace, from 4.72 lakh in 2001 to more than seven lakhs in the 2011 census report (District Census Handbook 2011 ). The research area was chosen because of its advantageous position as a center for trade and business, tourist activity, population expansion, the hub of employment, and essential supply to the entire northeast region (Roy et al. 2022a , b ). Because of this, the city's air quality suffers as traffic levels increase, which in turn creates more air pollution (The Statesman 2018b ; Roy and Singha 2020 ). Being the connecting point between the north-eastern states and the rest of India, Siliguri has earned the titles like "Gateway of North-east India" and "The chicken's neck" (Bose and Chowdhury 2020 ; Roy et al. 2022b ). This rapidly urbanized city extends over the Jalpaiguri and Darjeeling districts of West Bengal. The city of Siliguri is located between the coordinates of 26°39′57.88" and 26°46′19.03"N and 88°25′16.47" and 88°26′53.62"E. In Siliguri, air pollution is a serious problem that is mostly caused by the city's fast urbanization and population expansion, which is also related to enormous spikes in the vehicle population plying on the road (CDP Report Siliguri 2015 ; CRCAP Report 2018 ; The Telegraph 2018 ; The Statesman 2018a ; Roy and Singha 2020 ). According to experts, the vast majority of diesel-powered vehicles operating in the city are the prime cause of ambient air pollution in the city (CDP Report Siliguri 2015 ; The Telegraph 2018 ; The Statesman 2018b ; Roy and Singha 2020 ). Along with it, extensive road construction and poor road dust management procedures are causing particle pollution levels in the city to climb (Fig. 1 ).

The geographic location of Siliguri city showing the continuous ambient air quality monitoring (CAAQM) station situated in Babupara (Ward 32)

To ascertain the connection between the air pollutants and meteorological factors of Siliguri city, all the data were collected and analyzed based on the data and information accessible from the Central Pollution Control Board (CPCB) and West Bengal Pollution Control Board (WBPCB). There is just one continuous ambient air quality monitoring station in Siliguri at present, and it is situated in Babupara (Ward 32), near Tinbatti more. The daily mean (24 h average) concentration of seven air pollutants, i.e., Particulate Matter (PM 2.5 , PM 10 ), Nitrogen Dioxide (NO 2 ), Sulphur Dioxide (SO 2 ), Carbon Monoxide (CO), Ozone (O 3 ), and Ammonia (NH 3 ), as well as four meteorological factors, i.e., Temperature, Relative humidity, Rainfall as well as Wind speed, have been obtained from the CPCB online portal for data acquisition from March 2018 to September 2022. CPCB implements stringent processes for sampling, analysis, and calibration in order to deliver data quality assurance with quality control programs.

Methodology

Initially, descriptive statistics were computed for both meteorological parameters and air pollutants. To find patterns in the concentration of air pollutants during the same time period, the Mann–Kendall trend analysis has been used. Tukey's HSD multiple comparisons were used to test for seasonal changes in air pollution concentration throughout the years at a 5% threshold of significance. The Spearman correlation analysis was evaluated to identify any associations between the air pollutants and the meteorological factors. Seasonal classifications used in this study, i.e., pre-monsoon (March–May), monsoon (June–September), post-monsoon (October–November), and winter (December–February), were taken from the study conducted by Sivaprasad and Babu ( 2014 ) and Dutta and Gupta ( 2021 ). Finally, in order to investigate the possible effects of temp, RH, RF, and WS on air pollutants' concentration, multiple linear as well as non-linear models were used. Yin et al. ( 2016 ) also considered only two parameters, i.e., Temp and RH as predictive factors to determine the daily PM concentration in Beijing, China.

Here, β 0 = model constant β 1 , β 2 , β 3 , β 4 , β 5 , β 6 , β 7 and β 8 = model parameters, $\hat{Y}$ = dependent variables (pollutants), and $X_1 ,X_2 , X_3 , X_4$ = independent variables (Temp, RH, RF, and WS, respectively). The overall methodology is illustrated in Fig. 2 through a flow chart for clear and comprehensive understanding.

Methodological flow chart adopted for the present study

Results and discussion

The outcomes of this study, together with discussions of the criterion air pollutants as well as meteorological parameters, are presented in the following sub-sections.

Descriptive statistics

The data regarding air pollutants, as well as meteorological parameters in Siliguri city ranging from March 2018 to September 2022, are summarized in Table 1 , Fig. 3 . Overall, the mean intensity of air pollutants, i.e., PM 2.5 , PM 10, NO 2, SO 2, CO, O 3, and NH 3 ranged from 6.02 to 290.38 µg/m3, 14.76 to 393.80 µg/m3, 2.16 to 113.31 µg/m3, 1.35 to 60.89 µg/m3, 0.1 to 2.04 mg/m3, 6.10 to 75.61 µg/m3, 1.06 to 166.41 µg/m3, respectively, from 2018 to 2022. Furthermore, meteorological factors, i.e., temp, RH, RF, and WS, ranged from 11.55 to 33.38 °C, 36.29 to 97.77%, 0.00 to 2.19 mm, 0.21 to 12.55 m/s, respectively.

Trends of concentration and descriptive statistics of criterion air pollutants as well as meteorological parameters from March 2018 till September 2022

Trends of air pollution levels and meteorological parameters

The trends of all seven air pollutants, as well as meteorological variables considered for this study (Temp, RH, RF, and WS) from March 2018 to September 2022 in Siliguri, are illustrated in Fig. 3 . There was a noticeable tendency of fluctuations in the concentrations of air pollutants and in the meteorological parameters during the period of the research (Fig. 3 ). Mann–Kendall (MK) trend analysis was used to look at whether or not the concentration of air pollutants has a trend (increasing or decreasing) over time. The final findings of the MK trend test are shown in Table 2 . Here, the alpha and Z values denote the statistical significance of the rising or falling trend in air pollution, while the tau value indicates the direction of the trend. Sen's slope's value reveals how quickly the trend is rising or falling on a daily basis.

Since the estimated p -values were less than the significance threshold of alpha = 0.05 and the null hypothesis of the Mann–Kendall analysis is that there was no trend in the concentration of air pollutants during the relevant period of the study, the research rejected the null hypothesis (Except NO2) and accepted the alternative hypothesis. With the exception of NO 2 and O 3 , all air pollutants had statistically significant trends at a 99% confidence level. At a 95% confidence level, the trend of O 3 is significant, but the trend of NO 2 is not significant. It is evident from Table 2 , Fig. 4 , that every pollutant had a trend over time. All of the pollutants, with the exception of NH 3 , exhibit a declining trend over time. The trend of NH 3 has been rising over the sands of time.

Area trend diagram showing the concentration of air pollutants in Siliguri city

Comparison of regression models

Multiple linear regression (MLR) and non-linear regression (MNLR) were compared to identify the possible impacts of meteorological factors (Temp, RH, RF, and WS) on the concentrations of air pollutants (See Table 3 ). The Coefficient of determination ( R 2 ) of both models has been calculated in order to evaluate their efficacy. Table 3 compares the two models and demonstrates that the multiple non-linear regression (MNLR) model performed slightly better than the multiple linear regression (MLR) model. Considering all four meteorological factors as explanatory variables in the model, variations in PM 2.5 concentration are best explained by the meteorological parameters ( R 2 = 0.65 and 0.67, respectively), followed by PM 10 ( R 2 = 0.61 and 0.62, respectively), while variations in SO 2 concentration are least explained by temperature, relative humidity, rainfall, as well as wind speed ( R 2 = 0.09 and 0.11, respectively). Yin et al. ( 2016 ) demonstrated superior MNLR model performance in explaining PM 2.5 concentration in connection to meteorological factors in China. For CO, O 3 , and NH 3 , however, no model did well enough. However, NO 2 shows a significant link with meteorological factors than they do. As per Table 3 , R 2 values for MLR and MNLR vary from 0.09 to 0.65 and 0.11 to 0.67, respectively. Wise and Comrie (2005) showed coefficient of determination ( R 2 ) values between 0.1 and 0.5 when modeling the impacts of meteorological data on particulate matter concentration in the United States.

Seasonal fluctuations of air pollutants concentrations

A one-way between-groups ANOVA was carried out to ascertain if there is a variation in the concentrations of air pollutants between the seasons (pre-monsoon, monsoon, post-monsoon, as well as winter). The mean difference in the amounts of pollutants throughout the seasons was then shown using Tuckey's HSD post-hoc multiple comparisons (see Table 4 ).

The concentration of each and every pollutant varied across the four seasons, and this variation was statistically significant at the 0.05 level. The seasonal concentrations of PM 2.5 and PM 10 over the years showed significant maximum differences (mean score) according to Tukey's HSD multiple comparison tests (See Table 4 ). In contrast, NO 2 , CO, O 3 , and NH 3 show moderate variation across seasons, whereas SO 2 shows negligible variation. It is important to note that the difference in concentration between pre- and post-monsoon for PM 2.5 and SO 2 and between monsoon and post-monsoon for NH 3 is not statistically significant at all (see Table 4 ). Although the seasonal concentration varies depending on the various pollutants, the monsoon season depicts the minimum concentration of air pollutants, and the winter portrays the maximum (see Table 4 ). When evaluating seasonal fluctuations of particulate matter, it is obvious that, for all years, the mean concentrations of PM 2.5 and PM 10 were much greater in the winter than they were during the monsoon seasons (Table 4 ). When compared to monsoon, the amounts of fine particulate matter increased every year during post-monsoon season. Winter and pre-monsoon CO, NO 2 , SO 2 , and NH 3 concentrations were greater than the monsoon and post-monsoon seasons over the years, notwithstanding the little changes. O 3 levels during pre-monsoon were consistently greater than during other times of the year. In addition to seasonal changes, meteorological variables primarily contribute to the causes of the seasonal swings in the pollutants level (Manju et al. 2018 ). In addition, winter pollution levels may be higher due to road dust, vehicle exhaust during the dry winter, and urban construction activities. On the other hand, the moist deposition of particles during the monsoon season is linked to lower concentration levels in that season (Roy and Singha 2020 ; Biswas et al. 2020 ).

Association between changes in meteorological parameters and the levels of air pollutants concentration

The relationship between seasonal fluctuations in air pollution concentrations and meteorological factors, i.e., Temp, RH, RF, and WS, were examined using a 5% significant level of spearman correlation analysis. The values of the correlation coefficient are shown in Table 5 . Insignificant relationships are denoted by a cross ( X ) in Fig. 5 , which shows the season-wise correlation matrix between air pollution concentration and meteorological factors. However, Fig. 6 , 7 , 8 , and 9 show graphical presentations of the correlation between air pollutant levels and seasonal Temp, RH, RF, and WS.

Correlation matrix showing the season wise ( a,b,c,d represents pre-monsoon, monsoon, post-monsoon and winter respectively) Spearman correlation coefficient values in between air pollutants and meteorological parameters

Correlation plot between season wise air pollutants concentration and temperature, a pre‐monsoon (March–May), b monsoon (June–September), c post‐monsoon (October–November), d winter (December–February)

Correlation plot between season wise air pollutants concentration and relative humidity, a pre‐monsoon (March–May), b monsoon (June–September), c post‐monsoon (October–November), d winter (December–February)

Correlation plot between season wise air pollutants concentration and rainfall, a pre‐monsoon (March–May), b monsoon (June–September), c post‐monsoon (October–November), d winter (December–February)

Correlation plot between season wise air pollutants concentration and wind speed, a pre‐monsoon (March–May), b monsoon (June–September), c post‐monsoon (October–November), d winter (December–February)

Table 5 shows that during the post-monsoon and monsoon seasons, PM 2.5 and PM 10 had strong negative as well as moderate positive correlations associated with atmospheric temperature, whereas during the pre-monsoon as well as winter seasons, there was negligible or no correlation (Figs. 5 , 6 ). The meteorological features of the monsoon season and their interactions with the particulate matter may be used to explain the positive link between air temperature and particulate matter (PM 2.5 and PM 10 ) during the monsoon season. Giri et al. ( 2008 ) and Kayes et al. ( 2019 ) in their study made comparable observations. They said that since the particulate matter is mostly composed of soil or road dust, it quickly settles to the ground after absorbing water vapor from the atmosphere. Furthermore, this research shows that when compared to other times of the year, PM 2.5 and PM 10 concentrations are at their lowest during the monsoon season. The monsoon season of West Bengal is characterized by an intense downpours, gusty winds, and warm temperatures (Chatterjee et al. 2016 ). Although airborne particles may settle to the ground this time of year due to the abundance of precipitation, their concentration might be enhanced by drying up as a consequence of intense summer heat (Kayes et al. 2019 ). Another factor that may interact with the weather this time of year is the high frequency and intensity of winds. A significant link between fine particulate matter and temperature in the United States was also observed by Tai et al. ( 2010 ). It is noteworthy that among other pollutants, NO2, SO2, CO, O3 as well as NH3 resulted in a moderately negative correlation with temperature in post-monsoon season (Table 5 ).

Most air pollutants’ concentrations had a negative association with relative humidity throughout the seasons (Table 5 , Figs. 5 , 7 ). In particular, the concentration of fine and coarse particulate matter (PM 2.5 and PM 10 ) resulted in a negative association during the pre-monsoon, monsoon, and post-monsoon ( r = − 0.71, − 0.20, − 0.14 and − 0.72, − 0.33, − 0.29, respectively); however, there was little positive to no association during the winter ( r = 0.20, − 0.04). This is due to the fact that relative humidity affects the mobility of particles and might lead them to settle on the ground. As a result, the concentration of air contaminants decreases as relative humidity rises (Giri et al. 2008 ). Similarly, our research discovered that there was often a negative association between relative humidity and particulate matter concentration. However, a positive link between wintertime humidity and fine particulate matter (PM 2.5 ) suggests that the lack of precipitation and water vapor in the air may assist PM2.5's ventilation effects (Giri et al. 2008 ). Again, because of the favorable weather, the majority of construction work is done in the winter. O 3 and NH 3 were always negatively associated, with the exception of the monsoon season. This study's findings on the adverse relationship between O 3 and relative humidity and those of Swamy et al. ( 2012 ), Kumar et al. ( 2014 ), and Manju et al. ( 2018 ) are consistent. According to Ojha et al. ( 2016 ), the summer monsoon has greater CO and O 3 concentrations than other times of the year. Similar findings were also reported in China by Zhang et al. ( 2015 ), who made the case that the direction of the wind has a big impact on the amount of CO and O3 in the air. Ojha et al. ( 2016 ) said that these air pollutants might be carried from the African area to the Indian subcontinent by monsoon wind from the southwestern direction. As an outcome, it exhibits a positive link with humidity during the winter when it is feasible for vehicles to exhale more CO into the atmosphere.

The relationship between the concentration of pollutants and seasonal rainfall throughout time is shown in Figs. 5 , 8 , and Table 5 . Most air pollution concentrations had a negative link with rainfall throughout all seasons when it came to relative humidity (Table 5 ). For the pre-monsoon and post-monsoon seasons, a moderate to the low negative connection between rainfall and pollutants has been noted. However, research also demonstrates a practically weak negative association between pollutants and rainfall throughout the winter. PM2.5 and PM10 have the strongest reaction (in terms of negative correlation) with rainfall throughout the seasons among all the pollutants.

Table 5 and Figs. 5 , 9 show the relationship between wind speed and pollution concentration. Although NH 3 has a moderately positive correlation with wind speed during the monsoon and post-monsoon seasons ( r = 0.47, 0.46) and particulate matter (PM 2.5 and PM 10 ) has a moderately negative correlation with wind speed during the winter ( r = − 0.40, − 0.48), most air pollutants’ concentrations have a weak to negligible negative correlation with wind speed across all seasons. Since NH 3 is a thinner gas than air and tends to rise, it typically does not concentrate in low-lying locations and cannot be fully described by simple meteorological factors, according to research by Osada ( 2020 ). In their investigation on the level of particulate matter concentration in winter, Cichowicz et al. ( 2020 ) suggested that because of the low wind speed, particulate matter concentration rises throughout the winter. The wind speed slows throughout the winter months in Siliguri as well, increasing particle pollution.

It was necessary to look into the relationship between the concentration of seasonal ambient air pollutants and variations in meteorological parameters throughout time in the city of Siliguri due to the dearth of previous research on this particular area. As shown by other investigations, Siliguri city's air quality is a major problem. When looking at data from 2018 to 2022, the concentrations of all criterion air pollutants in Siliguri city have a decreasing trend; however, the concentration of NH 3 is rising. This could be attributable to COVID-19, the effects of which would be felt in reduced pollution levels in 2020 and 2021 as a result of the continuous lockdown associated with a substantial decrease in vehicle movement and other activities, such as construction works, which contributed to a significant decrease in criteria air pollutants. Although there were other air contaminants present, however, a negative correlation was shown between PM 2.5 and PM 10 and the four meteorological parameters. The study found that during the monsoon season, PM 2.5 , PM 10 , and temperature all have a positive relationship, in contrast to the negative relationships seen during the other seasons. This suggests that the high temperature and humidity experienced during this time contribute to the suspension of particulate matter. However, dry air with reduced humidity aggregates greater pollutants in urban environments; hence winter has the highest concentrations of air pollutants among the seasons. Although not the primary focus of this study, it is worth noting that particulate matters (PM 2.5 and PM 10 ) were found to be in infringement of the allowable limit established by the National Ambient Air Quality Standards in Siliguri city, posing a serious risk to the city's residents. The high quantities of PM 10 and PM 2.5 in the city's air are mostly attributable to air pollution, especially caused by diesel-powered automobiles. Compressed natural gas and electric cars are two examples of greener fuels that need to replace traditional options as quickly as feasible. There has to be an immediate increase in the number of air quality monitoring stations around the city. The renewal period for Pollution Under Control (PUC) certificates should be shortened to encourage more frequent renewal and reduce the number of cars owned by individual households. Meanwhile, construction sites must immediately adopt measures to cut down on dust-particle emissions. Enhancing the public transportation system should also be relevant to improving the efficiency of roads and services. The current study only used four meteorological parameters, including temperature, relative humidity, rainfall, and wind speed, to establish the associations, despite the fact that it is well-known that wind direction, solar radiations, and other factors can have an impact on the concentration of air pollutants. Future research should integrate more parameters to express this issue.

Data availability

The datasets used in the present study are publicly available from the official website of the Central Pollution Control Board (CPCB) and West Bengal Pollution Control Board (WBPCB).

Afghan FR, Habib H, Akhunzada NA, Wafa W, Shirzad MD, Sahak K, Hashimi SK, Mujeeb M, Wardak K, Ahmadzai MR (2022) Customization of GIS for spatial and temporal analyses of Air Quality Index trends in Kabul city. Model Earth Syst Environ 8(4):5097–5106. https://doi.org/10.1007/s40808-022-01396-5

Article Google Scholar

Agarwal A, Kaushik A, Kumar S, Mishra RK (2020) Comparative study on air quality status in Indian and Chinese cities before and during the COVID-19 lockdown period. Air Qual Atmos Health 13(10):1167–1178. https://doi.org/10.1007/s11869-020-00881-z

Anwar MN, Shabbir M, Tahir E, Iftikhar M, Saif H, Tahir A, Murtaza MA, Khokhar MF, Rehan M, Aghbashlo M, Tabatabaei M, Nizami AS (2021) Emerging challenges of air pollution and particulate matter in China, India, and Pakistan and mitigating solutions. J Hazardous Mater 416:125851. https://doi.org/10.1016/j.jhazmat.2021.125851

Bernard SM, Samet JM, Grambsch A, Ebi KL, Romieu I (2001) The potential impacts of climate variability and change on air pollution-related health effects in the United States. Environ Health Perspect 109(Suppl 2):199–209. https://doi.org/10.1289/ehp.109-1240667

Bhaskar BV, Mehta VM (2010) Atmospheric particulate pollutants and their relationship with meteorology in Ahmedabad. Aerosol Air Qual Res 10(4):301–315. https://doi.org/10.4209/aaqr.2009.10.0069

Bhunia P, Das P, Maiti R (2019) Meteorological drought study through SPI in three drought prone districts of West Bengal, India. Earth Syst Environ 4(1):43–55. https://doi.org/10.1007/s41748-019-00137-6

Biswas K, Chatterjee A, Chakraborty J (2020) Comparison of air pollutants between Kolkata and Siliguri, India, and its relationship to temperature change. J Geovisual Spat Anal. https://doi.org/10.1007/s41651-020-00065-4

Borge R, Requia WJ, Yagüe C, Jhun I, Koutrakis P (2019) Impact of weather changes on air quality and related mortality in Spain over a 25 year period [1993–2017]. Environ Int 133:105272. https://doi.org/10.1016/j.envint.2019.105272

Bose A, Chowdhury IR (2020) Monitoring and modeling of spatio-temporal urban expansion and land-use/land-cover change using markov chain model: a case study in Siliguri Metropolitan area, West Bengal, India. Model Earth Syst Environ 6(4):2235–2249. https://doi.org/10.1007/s40808-020-00842-6

Castelhano FJ, Pedroso AC, Cobelo I, Borge R, Roig HL, Adams M, Amini H, Koutrakis P, Réquia WJ (2022) The impact of long-term weather changes on air quality in Brazil. Atmos Environ 283:119182. https://doi.org/10.1016/j.atmosenv.2022.119182

CDP Report Siliguri (2015) City development plan for Siliguri 2041. Capacity building for urban development project, ministry of urban development. http://siligurismc.in/userfiles/file/siliguri-CDP-final-report-29April15.pdf

Chatterjee S, Khan A, Akbari H, Wang Y (2016) Monotonic trends in spatio-temporal distribution and concentration of monsoon precipitation (1901–2002), West Bengal, India. Atmos Res 182:54–75. https://doi.org/10.1016/j.atmosres.2016.07.010

Chatterji A (2021) Air pollution in Delhi: filling the policy gaps. Massach Undergr J Econ 17:1

Google Scholar

Chen Y, Wild O, Conibear L, Ran L, He J, Wang L, Wang Y (2020) Local characteristics of and exposure to fine particulate matter (PM2.5) in four Indian megacities. Atmos Environ 5:100052. https://doi.org/10.1016/j.aeaoa.2019.100052

Cichowicz R, Wielgosiński G, Fetter W (2020) Effect of wind speed on the level of particulate matter PM10 concentration in atmospheric air during winter season in vicinity of large combustion plant. J Atmos Chem 77(1–2):35–48. https://doi.org/10.1007/s10874-020-09401-w

CRCAP Report (2018) Climate resilient city action plan—Siliguri. Capacities. http://capacitiesindia.org/wp-content/uploads/2019/12/CRCAP_Siliguri_Sept2018.pdf

Dayan U, Levy I (2005) The influence of meteorological conditions and atmospheric circulation types on PM10 and visibility in Tel Aviv. J Appl Meteorol 44(5):606–619. https://doi.org/10.1175/jam2232.1

District Census Handbook, Darjiling (2011). Census of India—2011. West Bengal. https://censusindia.gov.in/2011census/dchb/1901_PART_B_DCHB_DARJILING.pdf

Doreswamy KSH, Yogesh KM, Gad I (2020) Forecasting air pollution particulate matter (PM2.5) using machine learning regression models. Proc Comput Sci 171:2057–2066. https://doi.org/10.1016/j.procs.2020.04.221

Dutta D, Gupta S (2021) Rising trend of air pollution and its decadal consequences on meteorology and thermal comfort over Gangetic West Bengal, India. In: Spatial Modeling and Assessment of Environmental Contaminants. Springer, Cham, pp 689–720

Elperin T, Fominykh A, Krasovitov B, Vikhansky A (2011) Effect of rain scavenging on altitudinal distribution of soluble gaseous pollutants in the atmosphere. Atmos Environ 45(14):2427–2433. https://doi.org/10.1016/j.atmosenv.2011.02.008

Gao C, Li S, Liu M, Zhang F, Achal V, Tu Y, Zhang S, Cai C (2021) Impact of the COVID-19 pandemic on air pollution in Chinese megacities from the perspective of traffic volume and meteorological factors. Sci Total Environ 773:145545. https://doi.org/10.1016/j.scitotenv.2021.145545

Giri D, Adhikary PR, Murthy VK (2008) The influence of meteorological conditions on PM10 concentrations in Kathmandu Valley

Glazener A, Khreis H (2019) Transforming our cities: best practices towards clean air and active transportation. Curr Environ Health Rep 6(1):22–37. https://doi.org/10.1007/s40572-019-0228-1

Gualtieri G, Toscano P, Crisci A, Di Lonardo S, Tartaglia M, Vagnoli C, Zaldei A, Gioli B (2015) Influence of road traffic, residential heating and meteorological conditions on PM10 concentrations during air pollution critical episodes. Environ Sci Pollut Res 22(23):19027–19038. https://doi.org/10.1007/s11356-015-5099-x

Gurjar B, Butler T, Lawrence M, Lelieveld J (2008) Evaluation of emissions and air quality in megacities. Atmos Environ 42(7):1593–1606. https://doi.org/10.1016/j.atmosenv.2007.10.048

Haddad K, Vizakos N (2020) Air quality pollutants and their relationship with meteorological variables in four suburbs of Greater Sydney, Australia. Air Qual Atmos Health 14(1):55–67. https://doi.org/10.1007/s11869-020-00913-8

Halder B, Bandyopadhyay J (2022) Air quality fluctuation monitoring and statistical data analysis during COVID-19 pandemic in Siliguri city, India. J Air Pollut Health. https://doi.org/10.18502/japh.v6i3.8232

He J, Gong S, Yu Y, Yu L, Wu L, Mao H, Song C, Zhao S, Liu H, Li X, Li R (2017) Air pollution characteristics and their relation to meteorological conditions during 2014–2015 in major Chinese cities. Environ Pollut 223:484–496. https://doi.org/10.1016/j.envpol.2017.01.050

Hoek G, Beelen R, Brunekreef B (2015) Land use regression models for outdoor air pollution. Exp Assess Environ Epidemiol. https://doi.org/10.1093/med/9780199378784.003.0013

Jain S, Sharma SK, Mandal TK, Saxena M (2018) Source apportionment of PM10 in Delhi, India using PCA/apcs, UNMIX and PMF. Particuology 37:107–118. https://doi.org/10.1016/j.partic.2017.05.009

Jain S, Sharma SK, Vijayan N, Mandal TK (2020) Seasonal characteristics of Aerosols (PM2.5 and PM10) and their source apportionment using PMF: A Four Year Study over Delhi, India. Environ Pollut 262:11437. https://doi.org/10.1016/j.envpol.2020.114337

Jat R, Gurjar BR (2021) Contribution of different source sectors and source regions of Indo-Gangetic Plain in India to PM2.5 pollution and its short-term health impacts during peak polluted winter. Atmos Pollut Res 12(4):89–100. https://doi.org/10.1016/j.apr.2021.02.016

Jayamurugan R, Kumaravel B, Palanivelraja S, Chockalingam MP (2013) Influence of temperature, relative humidity and seasonal variability on ambient air quality in a coastal urban area. Int J Atmos Sci 2013:1–7. https://doi.org/10.1155/2013/264046

Jhun I, Coull BA, Schwartz J, Hubbell B, Koutrakis P (2015) The impact of weather changes on air quality and health in the United States in 1994–2012. Environ Res Lett 10(8):084009. https://doi.org/10.1088/1748-9326/10/8/084009

Kan H, Chen R, Tong S (2012) Ambient air pollution, climate change, and population health in China. Environ Int 42:10–19. https://doi.org/10.1016/j.envint.2011.03.003

Kayes I, Shahriar SA, Hasan K, Akhter M, Kabir MM, Salam MA (2019) The relationships between meteorological parameters and air pollutants in an urban environment. Glob J Environ Sci Manag 5(3):265–278

Kulkarni GE, Muley AA, Deshmukh NK, Bhalchandra PU (2018) Autoregressive integrated moving average time series model for forecasting air pollution in Nanded City, Maharashtra, India. Model Earth Syst Environ 4(4):1435–1444. https://doi.org/10.1007/s40808-018-0493-2

Kumar A, Gurjar B (2019) Low-cost sensors for air quality monitoring in developing countries—a critical view. Asian J Water Environ Pollut 16(2):65–70. https://doi.org/10.3233/ajw190021

Kumar A, Singh BP, Punia M, Singh D, Kumar K, Jain VK (2014) Determination of volatile organic compounds and associated health risk assessment in residential homes and hostels within an academic institute. New Delhi Indoor Air 24(5):474–483. https://doi.org/10.1111/ina.12096

Kundu SK (2020) Rainfall trend analysis in the lower gangetic plain: a study on selected rainfall stations of Maldah District, West Bengal, India. Curr Sci 119(6):1031. https://doi.org/10.18520/cs/v119/i6/1031-1038

Leung KM, Yeung KW, You J, Choi K, Zhang X, Smith R, Zhou G, Yung MM, Arias-Barreiro C, An Y, Burket SR, Dwyer R, Goodkin N, Hii YS, Hoang T, Humphrey C, Iwai CB, Jeong S, Juhel G, Brooks BW (2020) Toward sustainable environmental quality: priority research questions for Asia. Environ Toxicol Chem 39(8):1485–1505. https://doi.org/10.1002/etc.4788

Li L, Qian J, Ou CQ, Zhou YX, Guo C, Guo Y (2014) Spatial and temporal analysis of Air Pollution Index and its timescale-dependent relationship with meteorological factors in Guangzhou, China, 2001–2011. Environ Pollut 190:75–81. https://doi.org/10.1016/j.envpol.2014.03.020

Li B, Shi XF, Liu YP, Lu L, Wang GL, Thapa S, Sun XZ, Fu DL, Wang K, Qi H (2020) Long-term characteristics of criteria air pollutants in megacities of Harbin-Changchun megalopolis, Northeast China: Spatiotemporal variations, source analysis, and meteorological effects. Environ Pollut 267:115441. https://doi.org/10.1016/j.envpol.2020.115441

Liang L, Wang Z, Li J (2019) The effect of urbanization on environmental pollution in rapidly developing urban agglomerations. J Clean Prod 237:117649. https://doi.org/10.1016/j.jclepro.2019.117649

Lin S, Wei D, Sun Y, Chen K, Yang L, Liu B, Huang Q, Paoliello MMB, Li H, Wu S (2020) Region-specific air pollutants and meteorological parameters influence COVID-19: a study from mainland China. Ecotoxicol Environ Saf 204:111035. https://doi.org/10.1016/j.ecoenv.2020.111035

Liu Y, Wang T (2020) Worsening urban ozone pollution in China from 2013 to 2017—part 1: the complex and varying roles of meteorology. Atmos Chem Phys 20(11):6305–6321. https://doi.org/10.5194/acp-20-6305-2020

Mage D, Ozolins G, Peterson P, Webster A, Orthofer R, Vandeweerd V, Gwynne M (1996) Urban air pollution in megacities of the world. Atmos Environ 30(5):681–686. https://doi.org/10.1016/1352-2310(95)00219-7

Manju A, Kalaiselvi K, Dhananjayan V, Palanivel M, Banupriya GS, Vidhya MH, Panjakumar K, Ravichandran B (2018) Spatio-seasonal variation in ambient air pollutants and influence of meteorological factors in Coimbatore, Southern India. Air Qual Atmos Health 11(10):1179–1189. https://doi.org/10.1007/s11869-018-0617-x

Mitoma H, Nagashima F, Kagawa S, Nansai K (2021) Critical supply chains for mitigating PM2.5 emission-related mortalities in India. Sci Rep. https://doi.org/10.1038/s41598-021-91438-2

Molina LT (2021) Introductory lecture: air quality in megacities. Faraday Discuss 226:9–52. https://doi.org/10.1039/d0fd00123f

Munsif R, Zubair M, Aziz A, Nadeem Zafar M (2021) Industrial air emission pollution: potential sources and sustainable mitigation. Environ Emissions. https://doi.org/10.5772/intechopen.93104

Ojha N, Pozzer A, Rauthe-Schöch A, Baker AK, Yoon J, Brenninkmeijer CAM, Lelieveld J (2016) Ozone and carbon monoxide over India during the summer monsoon: regional emissions and transport. Atmos Chem Phys 16(5):3013–3032. https://doi.org/10.5194/acp-16-3013-2016

Osada K (2020) Measurement report: short-term variation in ammonia concentrations in an urban area increased by mist evaporation and emissions from a forest canopy with bird droppings. Atmos Chem Phys 20(20):11941–11954. https://doi.org/10.5194/acp-20-11941-2020

Ouyang W, Guo B, Cai G, Li Q, Han S, Liu B, Liu X (2015) The washing effect of precipitation on particulate matter and the pollution dynamics of rainwater in downtown Beijing. Sci Total Environ 505:306–314. https://doi.org/10.1016/j.scitotenv.2014.09.062

Pachón J, Galvis B, Lombana O, Carmona L, Fajardo S, Rincón A, Meneses S, Chaparro R, Nedbor-Gross R, Henderson B (2018) Development and evaluation of a comprehensive atmospheric emission inventory for air quality modeling in the megacity of Bogotá. Atmosphere 9(2):49. https://doi.org/10.3390/atmos9020049

Pal S, Mahato S, Mandal I, Das P, Joshi P (2022) Evaluating aerosols concentration and air quality of Indian urban agglomerations over nationwide and regional lockdown. Atmos Pollut Res 13(10):101567. https://doi.org/10.1016/j.apr.2022.101567

Peng L, Zhao X, Tao Y, Mi S, Huang J, Zhang Q (2020) The effects of air pollution and meteorological factors on measles cases in Lanzhou, China. Environ Sci Pollut Res 27(12):13524–13533. https://doi.org/10.1007/s11356-020-07903-4

Piracha A, Chaudhary MT (2022) Urban air pollution, urban heat island and human health: a review of the literature. Sustainability 14(15):9234. https://doi.org/10.3390/su14159234

Rajput S, Arora K, Mathur R (2021) Level and extent of air pollution in Delhi. Urban Green Space, Health Economics and Air Pollution in Delhi, pp 59–88. https://doi.org/10.4324/9780429340581-4

Ramanathan V, Feng Y (2009) Air pollution, greenhouse gases and climate change: global and regional perspectives. Atmos Environ 43(1):37–50. https://doi.org/10.1016/j.atmosenv.2008.09.063

Roy S, Singha N (2021) Reduction in concentration of PM2.5 in India’s top most polluted cities: with special reference to post-lockdown period. Air Qual Atmos Health 14(5):715–723. https://doi.org/10.1007/s11869-020-00974-9

Roy S, Basak D, Bose A, Chowdhury IR (2022a) Citizens’ perception towards landfill exposure and its associated health effects: a PLS-SEM based modeling approach. Environ Monit Assess. https://doi.org/10.1007/s10661-022-10722-4

Roy S, Bose A, Majumadar S, Roy Chowdhury I, Abdo HG, Almohamad H, Al Dughairi AA (2022b) Evaluating urban environment quality (UEQ) for class-I Indian city: an integrated RS-GIS based exploratory spatial analysis. Geocarto Int. https://doi.org/10.1080/10106049.2022.2153932

Roy S, Singha N (2020) Analysis of ambient air quality based on exceedance factor and air quality index for Siliguri City, West Bengal. Curr World Environ. https://doi.org/10.12944/cwe.15.2.11

Seo J, Park DSR, Kim JY, Youn D, Lim YB, Kim Y (2018) Effects of meteorology and emissions on urban air quality: a quantitative statistical approach to long-term records (1999–2016) in Seoul, South Korea. Atmos Chem Phys 18(21):16121–16137. https://doi.org/10.5194/acp-18-16121-2018

Sharma P, Sharma P, Jain S, Kumar P (2013) An integrated statistical approach for evaluating the exceedence of criteria pollutants in the ambient air of megacity Delhi. Atmos Environ 70:7–17. https://doi.org/10.1016/j.atmosenv.2013.01.004

Shi Y, Bilal M, Ho HC, Omar A (2020) Urbanization and regional air pollution across South Asian developing countries—a nationwide land use regression for ambient PM2.5 assessment in Pakistan. Environ Pollut 266:15145. https://doi.org/10.1016/j.envpol.2020.115145

Sivaprasad P, Babu CA (2014) Seasonal variation and classification of aerosols over an inland station in India. Meteorol Appl 21(2):241–248

Song C, Wu L, Xie Y, He J, Chen X, Wang T, Lin Y, Jin T, Wang A, Liu Y, Dai Q, Liu B, Wang Y-N, Mao H (2017) Air pollution in China: status and spatiotemporal variations. Environ Pollut 227:334–347. https://doi.org/10.1016/j.envpol.2017.04.075

Srivastava AK (2022) Air pollution: facts, causes, and impacts. Asian Atmos Pollut. https://doi.org/10.1016/b978-0-12-816693-2.00020-2

Suhaimi NF, Jalaludin J, Latif MT (2020) Demystifying a possible relationship between COVID-19, air quality and meteorological factors: evidence from Kuala Lumpur, Malaysia. Aerosol Air Qual Res 20(7):1520–1529. https://doi.org/10.4209/aaqr.2020.05.0218

Surya B, Ahmad DNA, Sakti HH, Sahban H (2020) Land use change, spatial interaction, and sustainable development in the metropolitan urban areas, South Sulawesi Province, Indonesia. Land 9(3):95. https://doi.org/10.3390/land9030095

Swamy Y, Venkanna R, Nikhil G, Chitanya D, Sinha P, Ramakrishna M, Rao A (2012) Impact of nitrogen oxides, volatile organic compounds and black carbon on atmospheric ozone levels at a semi arid urban Site in Hyderabad. Aerosol Air Qual Res 12(4):662–671. https://doi.org/10.4209/aaqr.2012.01.0019

Tai APK, Mickley LJ, Jacob DJ (2010) Correlations between fine particulate matter (PM2.5) and meteorological variables in the United States: Implications for the sensitivity of PM2.5 to climate change. Atmos Environ 44(32):3976–3984. https://doi.org/10.1016/j.atmosenv.2010.06.060

The Statesman (2018a) Call to solve Siliguri pollution problems. https://www.thestatesman.com/cities/call-solve-siliguri-pollution-problems-1502623872.html

The Statesman (2018b) Man seeks answers on Siliguri pollution. https://www.thestatesman.com/cities/man-seeks-answers-siliguri-pollution-1502606581.html

The Telegraph (2018) Siliguri smokes into pollution top heap. https://www.telegraphindia.com/west-bengal/siliguri-smokes-into-pollution-topheap/cid/1453963

Tiwari S, Bisht D, Srivastava A, Pipal A, Taneja A, Srivastava M, Attri S (2014) Variability in atmospheric particulates and meteorological effects on their mass concentrations over Delhi, India. Atmos Res 145–146:45–56. https://doi.org/10.1016/j.atmosres.2014.03.027

Wei G, Zhang Z, Ouyang X, Shen Y, Jiang S, Liu B, He BJ (2021) Delineating the spatial–temporal variation of air pollution with urbanization in the Belt and Road Initiative area. Environ Impact Assess Rev 91:106646. https://doi.org/10.1016/j.eiar.2021.106646

Wen Y, Zhang S, Zhang J, Bao S, Wu X, Yang D, Wu Y (2020) Mapping dynamic road emissions for a megacity by using open-access traffic congestion index data. Appl Energy 260:114357. https://doi.org/10.1016/j.apenergy.2019.114357

Whiteman CD, Hoch SW, Horel JD, Charland A (2014) Relationship between particulate air pollution and meteorological variables in Utah’s Salt Lake Valley. Atmos Environ 94:742–753. https://doi.org/10.1016/j.atmosenv.2014.06.012

World Health Organization (WHO) (2016) Ambient air pollution: a global assessment of exposure and burden of disease

World Health Organization (WHO) (2019) World air quality report

World Health Organization (2021) WHO global air quality guidelines: particulate matter (PM2.5 and PM10), ozone, nitrogen dioxide, sulfur dioxide and carbon monoxide. World Health Organization

Yin Q, Wang J, Hu M, Wong H (2016) Estimation of daily PM2. 5 concentration and its relationship with meteorological conditions in Beijing. J Environ Sci 48:161–168

Zhang H, Wang Y, Hu J, Ying Q, Hu XM (2015) Relationships between meteorological parameters and criteria air pollutants in three megacities in China. Environ Res 140:242–254

Zhang Y, Wang L, Tang Z, Zhang K, Wang T (2022) Spatial effects of urban expansion on air pollution and eco-efficiency: evidence from multisource remote sensing and statistical data in China. J Clean Prod 367:132973. https://doi.org/10.1016/j.jclepro.2022.132973

Download references

Acknowledgements

First, the authors would like to express their sincere gratitude to the Department of Geography and Applied Geography, University of North Bengal, for providing the opportunity to conduct the research work. Besides, the authors would like to express their sincere gratitude to the Central Pollution Control Board (CPCB) and West Bengal Pollution Control Board (WBPCB) for the data, without which it is not possible to find such a novel result. This research paper was completed during the tenure of the UGC-JRF period. Besides, the authors would also like to thank Mr. Subham Roy and Mr. Suranjan Majumder (co-researchers) for supporting the ideas of this research. Lastly, the author wants to express their sincere gratitude to the anonymous referees and the editor in chief Md. Nazrul Islam for their insightful suggestions and comments, which greatly helped in the improvement of the earlier version of the manuscript.

Author information

Authors and affiliations.

Department of Geography and Applied Geography, University of North Bengal, Siliguri, West Bengal, 734013, India

Arghadeep Bose & Indrajit Roy Chowdhury

You can also search for this author in PubMed Google Scholar

Contributions

AB: Writing—Original Draft, Review and Editing, Conceptualization, Methodology, Software, Validation, Supervision. IRC: Review and Editing.

Corresponding author

Correspondence to Arghadeep Bose .

Ethics declarations

Conflict of interest.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Consent to publish

Both authors have read the manuscript carefully before submitting it and given consent to submit.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Bose, A., Roy Chowdhury, I. Investigating the association between air pollutants’ concentration and meteorological parameters in a rapidly growing urban center of West Bengal, India: a statistical modeling-based approach. Model. Earth Syst. Environ. 9 , 2877–2892 (2023). https://doi.org/10.1007/s40808-022-01670-6

Download citation

Received : 23 October 2022

Accepted : 28 December 2022

Published : 05 January 2023

Issue Date : June 2023

DOI : https://doi.org/10.1007/s40808-022-01670-6

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Air pollutants
Seasonal fluctuations
Particulate matter 2.5
Particulate matter 10
Relative humidity
Regression analysis
Find a journal
Publish with us
Track your research

Open access
Published: 12 February 2019

Visualization of real-time monitoring datagraphic of urban environmental quality

Pengyu Chen ORCID: orcid.org/0000-0003-2383-2881 1

EURASIP Journal on Image and Video Processing volume 2019 , Article number: 42 ( 2019 ) Cite this article

9647 Accesses

42 Citations

2 Altmetric

Metrics details

Quality of urban environment directly affects people health, and it is important to understand the real-time status of urban air quality. Air quality monitoring, data analysis, and visualization can grasp the concentration data of air pollutants in cities. In view of the current air quality monitoring using digital displays, it is difficult for users to intuitively determine the air pollution level with unsatisfied interaction mode of the data query. Using the real-time monitoring data of 23 observation points in Beijing, the work based on Google Earth applied Keyhole Markup Language (KML) for the visualization of air monitoring data. The interactive query makes it easier for users to query air quality, and gradually varied color can visually highlight the air quality level. Visualization of data has stronger expression (more images and more intuitive) than the original data table, which is beneficial for further analysis of data.

1 Introduction

The quality of the urban environment is related to the health of urban residents. Under global warming and urbanization, population concentration is getting higher and higher, with urban environmental quality becoming prominent. The coordinated development of human settlements and urbanization has become the focus in the world. At 2017 World Climate Conference in Copenhagen, members, focusing on energy conservation and emission reduction as well as the ecology and low carbon development, signed the Copenhagen Accord [ 1 ].

At present, most cities have established the complete monitoring system for environmental quality to measure six parameters (PM2.5, PM10, SO2, NO2, CO, O3) of air quality index (AQI), wind direction, noise, temperature and humidity, negative oxygen ions, and light, wherein AQI is the focus of real-time monitoring, especially for PM2.5 and PM10, because respirable particulate matter is a typical carcinogen. Urban air pollution leads to an increase in cancer incidence, and people can reduce outdoor activities due to air pollution. Real-time monitoring of air quality has the advantages such as guiding the travel of residents, regulating urban infrastructure traffic, strengthening medical care for respiratory diseases, and insisting environmental protection departments for reducing/stopping production of pollution-emitting enterprises.

As early as 1996, China formulated Ambient Air Quality Standard , which details the classification of urban air quality, standard classification, major pollutants, and their concentration limits at various levels. At present, China’s industry is still a traditional energy consumption structure. For economic development, a large amount of petrochemical energy such as coal and oil is consumed every year, causing serious pollution to the air. In winter in the north, extreme foggy weather appears due to heating, which leads to school suspension, factory shutdowns, and traffic control. It is the basis for improving urban air quality and the guarantee for the health of the masses by integrating computer network, air quality sensing, and data visualization into an urban air quality monitoring system.

Advances in current sensor technology and Internet of Things technology have made it possible to monitor air quality in various areas of the city in real time. However, the large amount of data collected in real time brings inconvenience to analysis and processing. Explosively growing big data sometimes exceeds the processing power of the system, and existing data mining techniques only use the tip of the iceberg [ 2 ]. Data visualization can transform data into intuitive graphical images and provide interaction and analysis between the server and users, showing the valuable rules between complex and massive data [ 3 ]. Data visualization is used to integrate the high-bandwidth, high-speed, and large-capacity storage vision system with the computer system of powerful computing and logic judgment. The graphic visualization of real-time data of urban air quality displays the air quality data obtained by real-time monitoring, enabling users to efficiently capture hidden features and patterns in big data. Without efficient datagraphic visualization analysis, it is impossible to achieve good management of urban air quality.

Based on Google Earth, the work proposed a graphical visualization method for real-time monitoring data of urban air quality, which is helpful to grasp the change and development trends of urban air quality and control pollution. It has theoretical significance and practical value for environmental management and public services.

2 Methods for visualized analysis

2.1 visualized analysis of air quality data.

There are a number of visualization methods available for analysis and processing of air quality data with one or more visualization tools. Qu et al. [ 4 ] used S-shaped parallel coordinates, a weighting map, and a polar coordinate system embedded in circular pixel strips to analyze the fog and haze of Hong Kong. Li et al. [ 5 ] used a multi-dimensional view to analyze air quality and meteorological data. A correlation detection view is proposed to visualize the change of air quality. Liao et al. [ 6 ] used the networked visualized analysis system to monitor air quality data in Beijing, with Geographical Information System (GIS), parallel coordinate, and splattering. Li et al. [ 7 ] analyzed the air pollution data in Beijing and used a two-dimensional diagram for the pollution levels of areas. Visualization for data analysis is difficult for ordinary users to understand, because the two-dimensional diagram of the data often fails to meet the growing demands of air quality information systems. Data should be presented in an interesting and easy-to-understand format to deliver information to end users.

In 2010, Canadian researches Aaron Van Daniela and Randall Martin added the total amount of aerosols monitored by NASA (National Aeronautics and Space Administration) satellites and superimposed them on the vertical distribution of aerosols calculated by computer models to obtain the global image of PM2.5 concentration distribution. PM2.5 pollution in China is very serious, and Shandong Province in China is the most polluted area of PM2.5 in the world. There is a common flaw in most of these traditional data visualization methods—they separate the relationship between air quality and time and space and statically calculate air quality data in the past. It cannot display air quality dynamically in real time, with no good predictability of future air quality.

2.2 Google Earth-based visualization system

Google Earth is a virtual Earth software developed by Google Inc., which arranges satellite photos, aerial photography, and GIS on a three-dimensional model of the Earth. Users can view high-definition satellite images from the world for free through the client software of Google Earth. The satellite images, not from a single data source, are the data fusion of satellite imagery and aerial imagery.

Charts are represented in a fixed scene in traditional air quality visualization methods and cannot be dynamically updated based on changes in location and real-time data. However, to master the air quality of a single city or a group of cities, it is impossible to rely on static charts without user interaction. The air quality of a city exists under specific time and space, and the charts are meaningless if it is deviating from geographical location and observation time.

Using Google Earth to visualize urban air quality combines air quality information with time and space. It visualizes massive amounts of air quality data in an intuitive and vivid way, breaking traditional patterns of data, formulas, and charts that express air quality. The simulation on the 3D virtual Earth platform can obtain the dynamic change of air quality to enhance the authenticity of data visualization of air quality. It is easy for users to understand and conducive to providing decision support for public management and improving urban air quality.

Figure 1 shows the average PM2.5 air quality index of monthly and weekly hotlist in the traditional table. Figure 2 shows the design sketch of data visualization based on Google Earth. From the comparison, the use of Google Earth to achieve data visualization of air quality is more intuitive and provides more images.

Average PM2.5 air quality index of monthly and weekly hotlist in the traditional table

Design sketch of data visualization based on Google Earth

At present, many scholars have performed lots of useful work in this field. Environmental information is integrated into Google Earth [ 8 ], and visual dynamic playback of pollution distribution is implemented based on Google Earth [ 9 ]. Urban air pollution is investigated by Google Earth with detailed explanation [ 10 ]. Google Earth is applied to the simulation of urban air pollution spread. Based on the distribution of atmospheric pollution concentration, dangerous areas are drawn in the urban map of Google Earth, providing the basis for emergency decision-making [ 11 ]. The data of 3D scanning lidar is integrated into Google Earth to observe real-time atmospheric pollution. The visualization of information can quickly determine the location of pollution sources and assess which areas are seriously affected [ 12 ].

3 Data visualization of urban air quality

3.1 keyhole markup language (kml).

KML is a markup language based on XML standard, which uses a markup structure with nested elements and attributes to express geotagging. KML files are written according to the KML language, widely used in software such as Google Earth, Google Map, and Google Maps for mobile [ 13 ]. The basic structure of a KML file is shown in a KML code, which uses a yellow pushpin to mark a location on the surface of the Earth. Place mark is the most commonly used geographic feature in Google Earth.

KML files are used to exchange geographic data in Google Earth. File-based data exchange has been one of the main ways of software data exchange. KML files define a number of TAGS used to specify how geographic data is displayed. Geographic features that can be defined include locations, descriptions, overlays, paths, and polygons.

3.2 Public data of air quality

With the rapid development of Internet technology, local government agencies provide public data on air quality monitoring to protect the health of the people. USA, Britain, Australia, and Canada have operated public data sites. For example, the US AirNow website specializes in providing air quality data and its impact on human health, helping people to make healthy behaviors based on different air quality [ 14 ]. UK Air Quality Archive is the UK’s authoritative platform of air quality query. The Chinese government has also established numerous public data sites. However, regarding air pollution data, the public generally obtains data from the mass media. Since numerical data is difficult to understand, it is necessary to visualize the data.

The work studied Beijing, which is located at 39″ 26′ to 41″ 03′ north latitude and 115″ 25′ to 117″ 30′ east longitude on the northwestern edge of the North China Plain with a total area of 164.1 thousand square kilometers and an urban area of 1401 thousand square kilometers. The west, north, and northeast of Beijing are surrounded by mountains, and the southeast is a large plain that slowly slopes towards the Bohai Sea. The altitude of the Beijing Plain is 20 to 60 m, and the mountain is generally 1000 to 1500 m above the sea level.

The real-time monitoring data of air quality in the work comes from the Beijing Municipal Environmental Monitoring Center established in 1974. The center is the first-level station of national environmental monitoring that provides environmental quality monitoring of environmental factors (such as atmosphere, water, noise, soil, and ecology), the monitoring of pollution sources, and the emergency monitoring of sudden pollution accidents in Beijing. Figure 3 shows the distribution of environmental assessment points in 23 districts across Beijing. The website’s data is updated every half hour, and the data of urban air quality can be obtained by the Web Service interface. Table 1 shows the specific parameters and sample data of air quality data in the work. The data also includes other information such as weather forecast, wind direction, typhoon, and monitoring station. Only air pollution data is used in the work.

In Table 1 , AQI represents air quality index with no unit in the range of 0–500, O 3 /8 h the average concentration of ozone within 8 h, PM2.5 the particles with a diameter of less than 2.5 μm, and PM10 the particles with a diameter of less than 10 μm.

Distribution of 23 air quality observation points in Beijing

3.3 Visualization process of air quality data

To improve the intuitiveness of air quality data, the work implemented data visualization of air quality based on Google Earth and KML. The real-time monitoring data of air quality was obtained from the Beijing Municipal Environmental Monitoring Center, stored in the server with regular update. By developing and applying the KML Generator program, the Web Service was used to request air quality data from the server and visualize the data by Google Earth. The generated KML code and its URL were stored in the public folder of Dropbox. They were linked to Google Earth via KML network links that read air quality data in the server in real time [ 15 , 16 , 17 ]. Once the server is properly configured with the shared URL of KML files, users who have installed Google Earth can view KML files hosted in the public web server. Figure 4 shows the specific process of reading data and visualizing air quality data using KML files.

Process of reading air quality data and data visualization

The KML code for real-time visualization of air quality using KML network link is shown below. The data in the server is read periodically, with the display refreshed at a specified time interval [ 18 , 19 ].

3.4 KML Generator program

KML Generator program requests the server to read the air quality data through the longitude and latitude of each observation point. Figure 5 shows the structure of KML files generated after obtaining the data. Each observation point corresponds to a folder containing <name>, <description>, and six <Placemark> tags. <name> indicates the name of the observation point, <description> shows a brief description of the observation area, and six <Placemark> corresponds to six kinds of polluting gases (including names and short descriptions). <Style> label indicates the color of each contaminant. The transparency of color depends on the level of contamination in the observation area—the darker color means the higher level of contamination.

KML structure for visualization of the pollution level in the observation area

3.5 Visualization of air quality

Figure 6 shows the visualization of air pollution data using KML and Google Earth. In the distribution of AQI, red indicates the high concentration of air pollutants, and blue the low concentration. The air pollution concentration is low in the north and high in the south, which is consistent with the geographical environment of Beijing. Forested land and orchards are mainly in the northern mountain areas of Beijing, with less settlements, factories, and roads. The south-central part is a bustling and crowed urban area with dense personnel and vehicles, resulting in a significantly lower air quality than the north. For the distribution of specific pollutants such as PM2.5, PM10, SO2, and NO2, users can switch to the single-pollutant distribution mode by clicking the label. The visualization in Fig. 6 can only show the overall air quality. More specific visualization results can be obtained by the interactive query in Google Earth.

Average AQI heat map in Beijing

4 Results and discussion

Visual interactive query allows users to learn more specific air quality values, performing a series of selection operations through the interactive graphical interface to query. Users intuitively transmit the query without complex query statements. The data shorted in the server can be directly read and written into KML format recognized by Google Earth [ 20 , 21 , 22 ]. The template changes according to different needs, and the air quality data is displayed in Google Earth with various forms.

Figure 7 shows the real-time AQI query for the area around Shunyi New City observation point. Users simply click on the observation point indicated by the red arrow in the graph to pop up the query results. In general, the size of the display screen is limited. To make users focus on a certain detail while browsing Google Earth, animation or a pop-up window (used in the work) is required to attract users’ attention.

Interactive query of air quality

To show the dynamic change process of the AQI in the past 24 h, a histogram is used to combine the cylinders of different colors to represent the AQI value. The green cylinder indicates the air quality level is excellent, with an AQI value between 0 and 50. The yellow one indicates the air quality level is good, with an AQI value between 0 and 50. The deepest purple indicates the air quality is severely polluted, with an AQI value greater than 300. For each air quality level, the effects of air on human health and recommended actions are briefly introduced. In this way, the overall understanding of Fig. 6 can be gradually refined to understand the specific air quality of each area.

Figure 8 shows the variation of contaminants within 24 h—six pollutants, temperature, pressure, and humidity. The horizontal axis in the bar chart is the time axis, indicating the last 48 h from now. The vertical axis represents the parameter values. On the far right is the actual value range of the parameter value within 48 h, not the theoretical value range of the parameter. The higher the cylinder height, the closer the parameter value is to the maximum value. The colors of the cylinder further distinguish different parameter values. As the color changes, the corresponding parameter value gradually becomes large. It shows the changes in air quality over the past period of time, and a rough estimate can be made accordingly on the concentration of each pollutant over a period of time. After the analysis process of more data, the laws implied in the big data make the air quality data more valuable.

Multiple air quality parameters

5 Conclusions

With the development of social economy, a large amount of harmful substances generated by industrial production and automobile exhaust gas are discharged into the air, causing serious air pollution. It is the most serious environmental problem facing most industrial cities in the world. Taking Beijing as an example, the work used the real-time monitoring data of 23 observation points throughout the city to visualize the monitoring data in the Google Earth. Users can understand the air quality distribution from a macro perspective or obtain the specific air quality data by the interactive query to understand the time trends of air pollution, pollutants, and air quality levels in different regions at different times.

In the future work, it is necessary to study the cloud storage for the data of each observation point, the storage of more data in a lone time range, and the acquisition of complete data information for further analysis and utilization of data. On the other hand, the reasonable air quality prediction model with big data will be studied to predict the air quality, thus meeting the needs of the public.

Abbreviations

Air quality index

Geographical Information System

Keyhole Markup Language

National Aeronautics and Space Administration

Doniger D (2009) The Copenhagen Accord: a big step forward, NRDC Climate Center. Accessed 21 Dec 2009

Google Scholar

H. Mayer, J. Holst, D. Schindler, D. Ahrens, Evolution of the air pollution in SW Germany evaluated by the long-term air quality index LAQx. Atmos. Environ. 42 (20), 5071–5078 (2008)

Article Google Scholar

Sarkar, D. (2008). Lattice: multivariate data visualization with R. Springer Science & Business Media

H. Qu, W.Y. Chan, A. Xu, K.L. Chung, K.H. Lau, P. Guo, Visual analysis of the air pollution problem in Hong Kong. IEEE Trans. Vis. Comput. Graph. 13 (6), 1408–1415 (2007)

J. Li, Z. Xiao, H.Q. Zhao, Z.P. Meng, K. Zhang, Visual analytics of smogs in China. J. Vis. 19 (3), 461–474 (2016)

Z. Liao, Y. Peng, Y. Li, X. Liang, Y. Zhao. A web-based visual analytics system for air quality monitoring data, 2014 22nd International Conference on Geoinformatics, Kaohsiung, The Institute of Electrical and Electronics Engineers, INC. 1-6 (Piscataway, NJ, 2014)

H. Li, H. Fan, F. Mao, A visualization approach to air pollution data exploration—a case study of air quality index (PM2. 5) in Beijing, China. Atmosphere 7 (3), 35 (2016)

N. Kussul, A. Shelestov, S. Skakun, Grid and sensor web technologies for environmental monitoring. Earth Sci. Inf. 2 (1–2), 37–51 (2009)

R. Varatharajan, G. Manogaran, M.K. Priyan, V.E. Balaş, C. Barna, Visual analysis of geospatial habitat suitability model based on inverse distance weighting with paired comparison analysis. Multimed. Tools Appl. 77 (14), 17573–17593 (2018)

M.J. URBAN, in 2009 Portland GSA Annual Meeting . Using Google Earth to examine metropolitan air pollution (2009)

Y. CHU, D. LIANG, J. GUO, X. FENG, J. FANG, Y. CHEN, Z. ZHANG, Z. YIN, Analysis of urban air pollution diffusion simulation based on Google Earth. Ind. Saf. Environ. Prot. 40 (09), 36–39 (2014)

C. Chiang, J. Nee, S. Das, S. Sun, Y. Hsu, H. Chiang, et al., in AGU Fall Meeting Abstracts . Incorporation of 3-D scanning lidar data into Google Earth for real-time air pollution observation (2011)

M. Silberbauer, W. Geldenhuys, in Academic track of the 2008 Free and Open Source Software for Geospatial (FOSS4G) Conference, incorporating the GISSA 2008 Conference, Cape Town, South Africa . Using keyhole markup language to create a spatial interface to South African water resource data through Google Earth (2008)

T. LU, Application of AIRNow ambient air quality notification system of USA in Shanghai. Environ. Monit. Forewarning 3 (01), 4–7 (2011)

S. Tavani, P. Granado, A. Corradetti, et al., Building a virtual outcrop, extracting geological information from it, and sharing the results in Google Earth via OpenPlot and Photoscan: an example from the Khaviz Anticline (Iran). Comput. Geosci. 63 , 44–53 (2014)

A. Chen, G. Leptoukh, S. Kempler, C. Lynnes, A. Savtchenko, D. Nadeau, J. Farley, Visualization of A-Train vertical profiles using Google Earth. Comput. Geosci. 35 (2), 419–427 (2009)

A. Triantafyllou, A. Watlet, C. Bastin, Geolokit: an interactive tool for visualising and exploring geoscientific data in Google Earth. Int. J. Appl. Earth Obs. Geoinf. 62 , 39–46 (2017)

Q. Zhou, Multi-layer affective computing model based on emotional psychology. Electron. Commer. Res. 18 (1), 109–124 (2018)

Q. Zhou, J. Luo, The study on evaluation method of urban network security in the big data era. Intell. Autom. Soft Comput. 24 (1), 133–138 (2018)

M. Richards, M. Ghanem, M. Osmond, Y. Guo, J. Hassard, Grid-based analysis of air pollution data. Ecol. Model. 194 (1–3), 274–286 (2006)

D. Christin, A. Reinhardt, S.S. Kanhere, M. Hollick, A survey on privacy in mobile participatory sensing applications. J. Syst. Softw. 84 (11), 1928–1946 (2011)

E. Kanjo, Noisespy: a real-time mobile phone platform for urban noise monitoring and mapping. Mob. Netw. Appl. 15 (4), 562–574 (2010)

Download references

Acknowledgements

Thank for the editor and reviewers.

No funding was received.

Availability of data and materials

Data will not be shared; reason for not sharing the data and materials is that the work submitted for review is not completed. The research is still ongoing, and those data and materials are still required by my team for further investigations.

Author information

Authors and affiliations.

Northwest University, Xi’an, 710127, China

Pengyu Chen

You can also search for this author in PubMed Google Scholar

Contributions

PC designed the research, analyzed the data, and wrote and edited the manuscript. The author read and approved the final manuscript.

Corresponding author

Correspondence to Pengyu Chen .

Ethics declarations

Ethics approval and consent to participate.

I approved.

Consent for publication

Competing interests.

The author declares that he has no competing interests. He has seen the manuscript and approved its submission to the journal. The author confirmed that the content of the manuscript has not been published or submitted for publication elsewhere.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License ( http://creativecommons.org/licenses/by/4.0/ ), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Reprints and permissions

About this article

Cite this article.

Chen, P. Visualization of real-time monitoring datagraphic of urban environmental quality. J Image Video Proc. 2019 , 42 (2019). https://doi.org/10.1186/s13640-019-0443-6

Download citation

Received : 18 September 2018

Accepted : 04 February 2019

Published : 12 February 2019

DOI : https://doi.org/10.1186/s13640-019-0443-6

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Data visualization
Google Earth
Air pollution

IMAGES

infographic air pollution with city highly air pollution illustration
Graphical representation of various sources and their % contribution
Air Pollution Causes Graph
Premium Vector
Air Pollution Infographic Template
Air pollution in world: real-time air quality index visual map

VIDEO

Air pollution processes and impacts (Arabic)
air pollution drawing (sources/types of air pollution
Tesscorn Cooper Environmental launches the new ADAPT Software
air pollution drawing||save environment poster
Plain Facts
Parameter for the true representation of organic pollution is

COMMENTS

Air Pollution
Air pollution is a health and environmental issue across all countries of the world but with large differences in severity. In the interactive map, we show death rates from air pollution across the world, measured as the number of deaths per 100,000 people in a given country or region.
Our Nation's Air 2021
Air Pollution Includes Gases and Particles. Air pollution consists of gas and particle contaminants that are present in the atmosphere. Gaseous pollutants include sulfur dioxide (SO 2), oxides of nitrogen (NO x), ozone (O 3), carbon monoxide (CO), volatile organic compounds (VOCs), and certain toxic air pollutants.Particle pollution (PM 2.5 and PM 10) includes a mixture of compounds that can ...
State of Global Air Report 2024
In the State of Global Air 2024 interactive app, you can explore, compare, and download data and graphics reflecting the latest air pollution levels and associated burden of disease for over 200 individual countries, territories, and regions, as well as track trends from 1990 to 2021. SoGA 2024 Report. 5.33 MB pdf.
Air pollution data portal
Air pollution data portal. From smog hanging over cities to smoke inside the home, air pollution poses a major threat to health across the globe. Almost all of the global population (99%) are exposed to air pollution levels that put them at increased risk for diseases including heart disease, stroke, chronic obstructive pulmonary disease ...
Health impacts of air pollution exposure from 1990 to 2019 in 43
Overall change of air quality parameters and morbidity and mortality estimates attributable to air pollution in Europe from 1990 to 2019. The average aPM 2.5 exposure of European countries was 20. ...
Visual analysis of global air pollution impact research: a bibliometric
The impact of air pollution is one of the hotspots attracting continuous scholarly attention, but the comprehensive statistical and visual analysis reviews are few. Employing the method of bibliometric analysis, this paper took the relevant literature from 1996 to April 2022 on the Web of Science as the research object. Through the methods of keyword co-occurrence analysis and burst analysis ...
AirVis: Visual Analytics of Air Pollution Propagation
Air pollution has become a serious public health problem for many cities around the world. To find the causes of air pollution, the propagation processes of air pollutants must be studied at a large spatial scale. However, the complex and dynamic wind fields lead to highly uncertain pollutant transportation. The state-of-the-art data mining approaches cannot fully support the extensive ...
Atmosphere
Visual exploration [9,10,11,12,13,14] of air pollution with spatio-temporal data is a solution that makes complex data understandable because graphical representation is relatively intuitive. Atmospheric particulate matter is a commonly used criterion to evaluate air quality [ 7 , 15 , 16 ].
Visual analysis of air pollution spatio-temporal patterns
Advances in air monitoring methods have made it possible to analyze large-scale air pollution phenomena. Mining potential air pollution information from large-scale air pollution data is an important issue in the current environmental field. Although direct data visualization provides an intuitive presentation, the method is less applicable in long-time domains with high temporal resolution ...
Statistical analysis for air pollution data
This chapter is designed as an introductory tutorial of common statistical analysis for air pollution data using R, a popular statistical software. ... calculation, and graphical representation of data from a sample. Descriptive statistics summarize data so we can understand the distribution, relation, or trend of data. However, we cannot use ...
Air Quality Trends
The blue band shows the distribution of air pollution levels among the trend sites, displaying the middle 80 percent. The black line represents the average among all the trend sites. Ninety percent of sites have concentrations below the top line, while ten percent of sites have concentrations below the bottom line. For each pollutant, the trend ...
Visualizing Air Quality Data
Air quality visualization - the graphical display of data - helps us understand the distribution of air pollutants in the atmosphere. This is hard to do just by looking at a modern air monitor equipment with its digital display. By combining real-time monitoring data with python programming, one can easily visualize air monitoring data.
Selecting Data Analytic and Modeling Methods to Support Air Pollution
Given the serious adverse health effects associated with many pollutants, and the inequitable distribution of these effects between socioeconomic groups, air pollution is often a focus of environmental justice (EJ) research. However, EJ analyses that aim to illuminate whether and how air pollution hazards are inequitably distributed may present a unique set of requirements for estimating ...
Visualization and Analysis of Air Pollution and Human Health Based on
The negative effects of air pollution on human health are well documented [1,2]. Harmful air pollutants, such as PM 2.5, PM 10, SO 2, and NO X, escaping into the environment through natural and human activities may adversely affect human health . Air pollution has acute and chronic effects on various systems and organs.
Graphical representation of various sources and their % contribution
Download scientific diagram | Graphical representation of various sources and their % contribution towards the emission of air pollutants a CO, b NO x , c SO 2 , and d CO 2 from publication ...
A Graphical Tool for Eliciting Knowledge of Air Pollution Sources
With tens of thousands of air quality monitoring stations installed in the world, this source of information has become the standard in air quality measuring. Air pollution becoming a growing concern for decades now, the need for an easy way to visualise pollution data arose. Extensive maps have been created to represent air pollution using data collected from monitoring stations, as well as ...
Graphical representation of air pollutants
Download scientific diagram | Graphical representation of air pollutants from publication: Air quality modeling for effective environmental management in Uttarakhand, India: A comparison of ...
Analysis of Air Pollution Data in India between 2015 and 2019
ABSTRACT India suffers from among the worst air pollution in the world. In response, a large government effort to increase air quality monitoring is underway. We present the first comprehensive analysis of government air quality observations from 2015-2019 for PM10, PM2.5, SO2, NO2 and O3 from the Central Pollution Control Board (CPCB) Continuous Ambient Air Quality Monitoring (CAAQM ...
The importance of the representation of air pollution emissions for the
Abstract. Aerosol particles can contribute to the Arctic amplification (AA) by direct and indirect radiative effects. Specifically, black carbon (BC) in the atmosphere, and when deposited on snow and sea ice, has a positive warming effect on the top-of-atmosphere (TOA) radiation balance during the polar day. Current climate models, however, are still struggling to reproduce Arctic aerosol ...
Visual Analytics of Air Pollutant Propagation Path and Pollution Source
Recently, controlling air pollution has become increasingly significant due to its impact on our health and daily lives. To prevent and control pollution, it is crucial to trace its source. Many researches have been developed for tracing the source of pollution.
Investigating the association between air pollutants' concentration and
The ambient air quality in a city is heavily influenced by meteorological conditions. The city of Siliguri, known as the "Gateway of Northeast India", is a major hotspot of air pollution in the Indian state of West Bengal. Yet almost no research has been done on the possible impacts of meteorological factors on criterion air pollutants in this rapidly growing urban area. From March 2018 to ...
Visualization of real-time monitoring datagraphic of urban
Quality of urban environment directly affects people health, and it is important to understand the real-time status of urban air quality. Air quality monitoring, data analysis, and visualization can grasp the concentration data of air pollutants in cities. In view of the current air quality monitoring using digital displays, it is difficult for users to intuitively determine the air pollution ...
PDF Air pollution attributed disease burden and economic growth in India
Air pollution-attributed disease burden estimates were already available in the public domain from the GBD study. However, the association of APADB with eco- ... Lorenz curve is the graphical (visual) representation of the concentration index. The Lorenz curve is a two-dimension (X and Y axis) scatter plot with a diagonal

Air Pollution

Indoor Air Pollution

Outdoor Air Pollution

Air pollution is one of the world's leading risk factors for death

Air pollution contributes to one in ten deaths globally

Air pollution is one of the leading risk factors for disease burden

Who is most affected by air pollution?

How are death rates from air pollution changing?

Interactive charts on air pollution

Cite this work

Reuse this work freely

OUR NATION'S AIR

Air Quality Trends Show Clean Air Progress

Air Pollution Includes Gases and Particles

Six Common Pollutants

Understanding Emission Sources Helps Control Air Pollution

Emission Inventories

Air Pollution Can Affect Our Health and Environment in Many Ways

Economic Strength with Cleaner Air

Criteria Pollutant Trends Show Clean Air Progress

Unhealthy Air Days Show Long-Term Improvement

Air Quality Index (AQI) Forecast

Air Quality in Nonattainment Areas Improves

Visibility Improves in Scenic Areas

Regional Haze Rule

Air Toxics Levels Trending Down

More Ways to Explore 2020 Air Quality

Air Quality Impacts From Wildfires

Fire and Smoke Map

State of Global Air Report 2024

Publications

Air pollution data portal

Interactive tools and resources

Visualization of national air quality standards

Clean Household Energy Solutions Toolkit (CHEST)

Household Multiple Emission Sources Model (HOMES)

Benefits of action to reduce household air pollution (BAR-HAP) tool

Performance Target (PT) Model

WHO global air quality guidelines: particulate matter (‎PM2.5 and PM10)‎, ozone, nitrogen dioxide, sulfur...

Personal interventions and risk communication on Air Pollution

Ambient air pollution: A global assessment of exposure and burden of disease

WHO Guidelines for indoor air quality: Household fuel combustion

Feedback for the new GHO portal

Health impacts of air pollution exposure from 1990 to 2019 in 43 European countries

Similar content being viewed by others

Disparities in air pollution attributable mortality in the US population by race/ethnicity and sociodemographic factors

Epidemiological trends and age-period-cohort effects on cardiovascular diseases burden attributable to ambient air pollution across BRICS

Global air pollution exposure and poverty

Sources of data

Definitions

Statistical analyses

Overall change of air quality parameters and morbidity and mortality estimates attributable to air pollution in Europe from 1990 to 2019

Progress of each country compared to Europe overall in terms of DALYs and aPM 2.5

Impact of economic and social factors on air pollution-related disease burden

Data availability

Acknowledgements

Author information

Contributions

Corresponding author

Ethics declarations

Additional information

Supplementary Information

About this article

Share this article

This article is cited by

Health effects of carbonaceous PM2.5 compounds from residential fuel combustion and road transport in Europe

A nationwide epidemiological study on the associations between ambient particulate matter and suicide in Finland

Air pollution, cardiorespiratory fitness and biomarkers of oxidative status and inflammation in the 4HAIE study

Addressing disparities in the global epidemiology of stroke

Quick links

Information

Initiatives

Article Menu

JSmol Viewer

1. Introduction

Appendix B. Data quality check of the U-Air data

Share and Cite

Article Metrics

Air Quality Trends - How to Interpret the Graphs

Visualizing Air Quality Data - Know through Graph of Air Pollution