graphical representation of water quality data

• DOI: 10.1111/J.1745-6584.1972.TB02912.X
• Corpus ID: 129629304

Graphical Interpretation of Water‐Quality Data

• A. Zaporozec
• Published 1 March 1972
• Environmental Science, Chemistry
• Ground Water

94 Citations

Some observations on the chemical classification of ground waters, gis-based hydrogeochemical analysis tools (quimet), interpretation of hydrochemical facies by factor analysis, application of multivariate statistical methods and water-quality index to evaluation of water quality in the kashkan river, ground water quality assessment using multi‐rectangular diagrams, significance of geographical, hydrogeological, and hydrogeochemical origin for the elemental composition of bottled german mineral waters, a review of gis-integrated statistical techniques for groundwater quality evaluation and protection, assessing ground water quality in winters of industrial zone, kota, rajasthan, major ion chemistry of environmental samples around sub-urban of chennai city, gis-based software platform for managing hydrogeochemical data, 11 references, a graphic procedure in the geochemical interpretation of water-analyses, digital computer methods for water‐quality dataa, geochemical patterns in coachella valley, duty's classification of natural waters and chemical composition of atmospheric precipitation in ussr: a review, hydrochemical facies and ground-water flow patterns in northern part of atlantic coastal plain, the interpretation of chemical water analysis by means of patterns, graphical methods for indicating the mineral character of natural waters, graphic representation of water analyses., a comparison of waters of mines and of hot springs, chemical relations of the oil-field waters in san joaquin valley, california, related papers.

Showing 1 through 3 of 0 Related Papers

Graphical interpretation of water quality data

• Published: June 1974
• Volume 3 , pages 217–236, ( 1974 )

Cite this article

• Jerome L. Mahloch 1  

145 Accesses

2 Citations

Explore all metrics

Management of our nation's water resources through planning and control of water pollution hinges on the availability and interpretation of water quality data on which to base management decisions. This paper is aimed at exploring graphical methods which allow rapid and informative analysis of water quality data.

The graphical methods presented in this paper fall into two main categories. The first category relates to graphical procedures which are developed as part of a statistical test, such as discriminant analysis. The second category involves approximation of multivariate water quality data into two dimensions. The methods for accomplishing the latter are canonical decomposition and high-dimensional plotting. Each of these methods is developed and used on an example set of data to demonstrate their utility. These methods seek to not only present a graphical representation of the data, but also to explain variations and interrelationships within the data itself.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Subscribe and save.

• Get 10 units per month
• Download Article/Chapter or eBook
• 1 Unit = 1 Article or 1 Chapter
• Cancel anytime

Price includes VAT (Russian Federation)

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

Similar content being viewed by others

A holistic review on the assessment of groundwater quality using multivariate statistical techniques

A practitioner’s guide for exploring water quality patterns using principal components analysis and procrustes.

A multivariate statistical approach to the integration of different land-uses, seasons, and water quality as water resources management tool

Andrews, D. F.: 1972, Biometrics 28 , 125.

Google Scholar  

Cooley, W. W. and Lohnes, P. R.: 1971, Multivariate Data Analysis , John Wiley and Sons, Inc. New York.

Gabriel, K. R.: 1971, Biometrika 58 , 453.

Gabriel, K. R.: 1972a, Candec Computer Program , Department of Statistics, The Hebrew University, Jerusalem.

Gabriel, K. R.: 1972b, J. App. Meteor. 11 , 1071.

Shannon, E. E. and Brezonik, P. L.: 1972, J. Sanit. Eng. Div., A.S.C.E. 98 , SAI, 37.

Download references

Author information

Authors and affiliations.

Dept. of Civil Engineering, Mississippi State University, 39762, Mississippi State, Miss., USA

Jerome L. Mahloch

You can also search for this author in PubMed   Google Scholar

Rights and permissions

Reprints and permissions

About this article

Mahloch, J.L. Graphical interpretation of water quality data. Water Air Soil Pollut 3 , 217–236 (1974). https://doi.org/10.1007/BF00166632

Download citation

Received : 06 December 1973

Revised : 25 February 1974

Issue Date : June 1974

DOI : https://doi.org/10.1007/BF00166632

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

• Water Quality
• Water Resource
• Water Pollution
• Graphical Representation
• Discriminant Analysis
• Find a journal
• Publish with us
• Track your research

Open Access is an initiative that aims to make scientific research freely available to all. To date our community has made over 100 million downloads. It’s based on principles of collaboration, unobstructed discovery, and, most importantly, scientific progression. As PhD students, we found it difficult to access the research we needed, so we decided to create a new Open Access publisher that levels the playing field for scientists across the world. How? By making research easy to access, and puts the academic needs of the researchers before the business interests of publishers.

We are a community of more than 103,000 authors and editors from 3,291 institutions spanning 160 countries, including Nobel Prize winners and some of the world’s most-cited researchers. Publishing on IntechOpen allows authors to earn citations and find new collaborators, meaning more people see your work not only from your own field of study, but from other related fields too.

Brief introduction to this section that descibes Open Access especially from an IntechOpen perspective

Want to get in touch? Contact our London head office or media team here

Our team is growing all the time, so we’re always on the lookout for smart people who want to help us reshape the world of scientific publishing.

Home > Books > Research and Practices in Water Quality

Validity and Errors in Water Quality Data — A Review

Submitted: 13 May 2014 Published: 09 September 2015

DOI: 10.5772/59059

Cite this chapter

There are two ways to cite this chapter:

From the Edited Volume

Research and Practices in Water Quality

Edited by Teang Shui Lee

To purchase hard copies of this book, please contact the representative in India: CBS Publishers & Distributors Pvt. Ltd. www.cbspd.com | [email protected]

Chapter metrics overview

2,841 Chapter Downloads

Impact of this chapter

Total Chapter Downloads on intechopen.com

Total Chapter Views on intechopen.com

Author Information

Innocent rangeti.

• Department of Environmental Health, Durban University of Technology, Durban, South Africa

Bloodless Dzwairo

• Institute for Water and Wastewater Technology, Durban University of Technology, Durban, South Africa

Graham J. Barratt

Fredrick a.o. otieno.

• DVC: Technology, Innovation and Partnerships, Durban University of Technology, Durban, South Africa

*Address all correspondence to: [email protected]

1. Introduction

While it is essential for every researcher to obtain data that is highly accurate, complete, representative and comparable, it is known that missing values, outliers and censored values are common characteristics of a water quality data-set. Random and systematic errors at various stages of a monitoring program tend to produce erroneous values, which complicates statistical analysis. For example, the central tendency statistics, particularly the mean and standard deviation, are distorted by a single grossly inaccurate data point. An error, which is initially identified and is later incorporated into a decision making tool, like a water quality index (WQI) or a model, could subsequently lead to costly consequences to humans and the environment.

Checking for erroneous and anomalous data points should be routine, and an initial stage of any data analysis study. However, distinguishing between a data-point and an error requires experience. For example, outliers may actually be results which might require statistical attention before a decision can be made to either discard or retain them. Human judgement, based on knowledge, experience and intuition thus continue to be important in assessing the integrity and validity of a given data-set. It is therefore essential for water resources practitioners to be knowledgeable regarding the identification and treatment of errors and anomalies in water quality data before undertaking an in-depth analysis.

On the other hand, although the advent of computers and various software have made it easy to analyse large amounts of data, lack of basic statistical knowledge could result in the application of an inappropriate technique. This could ultimately lead to wrong conclusions that are costly to humans and the environment [ 1 ]. Such necessitate the need for some basic understanding of data characteristics and statistics methods that are commonly applied in the water quality sector. This chapter, discusses common anomalies and errors in water quality data-sets, methods of their identification and treatment. Knowledge reviewed could assist with building appropriate and validated data-sets which might suit the statistical method under consideration for data analysis and/or modelling.

2. Data errors and anomalies

Referring to water quality studies, an error can be defined as a value that does not represent the true concentration of a variable such as turbidity. These may arise from both human and technical error during sample collection, preparation, analysis and recording of results [ 2 ]. Erroneous values can be recorded even where an organisation has a clearly defined monitoring protocol. If invalid values are subsequently combined with valid data, the integrity of the latter is also impaired [ 1 ]. Incorporating erroneous values into a management tool like a WQI or model, could result in wrong conclusions that might be costly to the environment or humans.

Data validation is a rigorous process of reviewing the quality of data. It assists in determining errors and anomalies that might need attention during analysis. Validation is crucial especially where a study depends on secondary data as it increases confidence in the integrity of the obtained data. Without such confidence, further data manipulation is fruitless [ 3 ]. Though data validation is usually performed by a quality control personnel in most organisations, it is important for any water resource practitioner to understand the common characteristics that may affect in-depth analysis of a water quality data-sets.

3. Visual scan

Among the common methods of assessing the integrity of a data-set is visual scan. This approach assists to identify values that are distinct and, which might require attention during statistical analysis and model building. The ability to visually assess the integrity of data depends on both the monitoring objectives and experience [ 4 ]. Transcription errors, erroneous values (e.g. a pH value of greater than 14, or a negative reading) and inaccurate sample information (e.g. units of mg/L for specific conductivity data) are common errors that can be easily noted by a visual scan. A major source of transcription errors is during data entry or when converting data from one format to another [ 5 , 6 ]. This is common when data is transferred from a manually recorded spreadsheet to a computer oriented format. The incorrect positioning of a decimal point during data entry is also a common transcription error [ 7 , 8 ].

A report by [ 7 ] suggested that transcription errors can be reduced by minimising the number of times that data is copied before a final report is compiled. [ 9 ] recommended the read-aloud technique as an effective way of reducing transcription errors. Data is printed and read-aloud by one individual, while the second individual simultaneously compares the spoken values with the ones on the original sheet. Even though the double data-entry method has been described as an effective method of reducing transcription errors, its main limitation is of being laborious [ 9 - 11 ]. [ 12 ], however, recommended slow and careful entry of results as an effective approach of reducing transcription errors.

While it might be easy to detect some of the erroneous values by a general visual scan, more subtle errors, for example outliers, may only be ascertained by statistical methods [ 13 ]. Censored values, missing values, seasonality, serial correlation and outliers are common characteristics in data-sets that need identification and treatment [ 14 ]. The following sections review the common characteristics in water quality data namely; outliers, missing values and censored values. Methods of their identification and treatment are discussed.

3.1. Outliers (extreme values)

The presence of values that are far smaller or larger than the usual results is a common feature of water quality data. An outlier is defined as a value that has a low probability of originating from the same statistical distribution as the rest of observations in a data-set [ 15 ]. Outlying values should be examined to ascertain if they are possibly erroneous. If erroneous, the value can be discarded or corrected, where possible. Extreme values may arise from an imprecise measurement tool, sample contamination, incorrect laboratory analysis technique, mistakes made during data transfer, incorrect statistical distribution assumption or a novel phenomenon, [ 15 , 16 ]. Since many ecological phenomena (e.g., floods, storms) are known to produce extreme values, their removal assumes that the phenomenon did not occur when actually it did. A decision must thus be made as to whether an outlying datum is an occasional value and an appropriate member of the data-set or whether it should be amended, or excluded from subsequent statistical analyses as it might introduce bias [ 1 ].

An outlying value should only be objectively rejected as erroneous after a statistical test indicates that it is not real or when it is desired to make the statistical testing more sensitive [ 17 ]. In figure 1 , for example, simple inspection might mean that the two spikes are erroneous, but in-depth analysis might correlate the spikes to very poor water quality for those two days, which would make the two observations valid. The model, however, does not pick the extreme values, which negatively affects the R 2 value, and ultimately the accuracy and usefulness of the model in predicting polymer dosage.

Data inspection during validation and treatment

Both observational (graphical) and statistical techniques have been applied to identify outliers. Among the common observational methods are the box-plots, time series, histogram, ranked data plots and normal probability plots [ 18 , 19 ]. These methods basically detect an outlier value by quantifying how far it lies from the other values. This could be the difference between the outlier and the mean of all points, between the outlier and the next closest value or between the outlier and the mean of the remaining values [ 20 ].

3.2. Box-plot

The box-plot, a graphical representation of data dispersion, is considered to be a simple observation method for screening outliers. It has been recommended as a primary exploratory tool of identifying outlying values in large data-sets (15). Since the technique basically uses the median value and not the mean, it poses a greater advantage by allowing data analysis disregarding its distribution. [ 21 ] and [ 22 ] categorised potential outliers using the box-plot as:

data points between 1.5 and 3 times the Inter Quantile Range (IQR) above the 75 th percentile or between 1.5 and 3 times the IQR below the 25 th percentile, and

data points that exceed 3 times the IQR above the 75 th percentile or exceed 3 times the IQR below the 25 th percentile.

The limitation of a box plot is that it is basically a descriptive method that does not allow for hypothesis testing, and thus cannot determine the significance of a potential outlier [ 15 ].

3.3. Normal probability plot

The probability plot method identifies outliers as values that do not closely fit a normal distribution curve. The points located along the probability plot line represent ‘normal’,observation, while those at the upper or lower extreme of the line, indicates the suspected outliers as depicted in Figure 2 .

Normal probability plot showing outliers

The approach assumes that if an extreme value is removed, the resulting population becomes normally distributed [ 21 ]. If, however, the data still does not appear normally distributed after the removal of outlying values, a researcher might have to consider normalising it by transformation techniques, such as using logarithms [ 21 , 23 ]. However, it should be highlighted that data transformation tends to shrink large values (see the two extreme values in Figure 1 , before transformation), thus suppressing their effect which might be of interest for further analysis [ 23 , 24 ]. Data should thus not be simply transformed for the sole purpose of eliminating or reducing the impact of outliers. Furthermore, since some data transformation techniques require non-negative values only (e.g. square root function) and a value greater than zero (e.g. logarithm function), transformation should not be considered as an automatic way of reducing the effect of outliers [ 23 ].

Since observational methods might fail to identify some of the subtle outliers, statistical tests may be performed to identify a data point as an outlier. However a decision still has to be made on whether to exclude or retain an outlying data point. The section below describes the common statistical test for identifying outliers.

3.4. Grubbs test

The Grubb’s test, also known as the Studentised Deviate test, compares outlying data points with the average and standard deviation of a data-set [ 25 - 27 ]. Before applying the Grubbs test, one should firstly verify that the data can be reasonably approximated by a normal distribution. The test detects and removes one outlier at a time until all are removed. The test is two sided as shown in the two equations below.

To test whether the maximum value is an outlier, the test:

To test whether the minimum value is an outlier, the test is:

Where X 1 or X n =the suspected single outlier (max or min)

s=standard deviation of the whole data set

The main limitation of Grubbs test is of being invalid when data assumes non-normal distribution [ 28 ]. Multiple iterations of data also tends to change the probabilities of detection. Grubbs test is only recommended for sample sizes of not more than six, since it frequently tags most of the points as outliers. It suffers from masking, which is failure to identify more than one outlier in a data-set [ 28 , 29 ]. For instance, for a data-set consisting of the following points; 3, 5, 7, 13, 15, 150, 153, the identification of 153 (maximum value) as an outlier might fail because it is not extreme with respect to the next highest value (150). However, it is clear that both values (150 and 153) are much higher than the rest of the data-set and could jointly be considered as outliers.

3.5. Dixon test

Dixon’s test is considered an effective technique of identifying an outlier in a data-set containing not more than 25 values [ 21 , 30 ]. It is based on the ratio of the ranges of a potential outlier to the range of the whole data set as shown in equation 1 [ 31 ]. The observations are arranged in ascending order and if the distance between the potential outlier to its nearest value (Q gap ) is large enough, relative to the range of all values (Q range ), the value is considered an outlier.

The calculated Q exp value is then compared to a critical Q-value (Qcrit) found in tables. If Q exp is greater than the suspect value, the suspected value can be characterised as an outlier. Since the Dixon test is based on ordered statistics, it tends to counter-act the normality assumption [ 15 ]. The test assumes that if the suspected outlier is removed, the data becomes normally distributed. However, Dixon’s test also suffers the masking effect when the population contains more than one outlier.

[ 32 ] recommended the use of multivariate techniques like Jackknife distance and Mahalanobis distance [ 33 , 34 ]. The strength of multivariate methods is on their ability to incorporation of the correlation or covariance between variables thus making them more correct as compared to univariate methods. [ 34 ] introduced the chi-square plot, which draws the empirical distribution function of the robust Mahalanobis distances against the chi-square distribution. A value that is out of distribution tail indicates that it is an outlier [ 33 ].

For an on-going study, an outlier can be ascertained by re-analysis of the sample, if still available and valid. [ 28 ] and [ 2 ] advised the practise of triplicate sampling as an effective method of verifying the unexpected results. When conducting a long-term study, researchers might consider re-sampling when almost similar conditions prevail again. Nevertheless, this option might not be feasible when carrying out a retrospective study since it generally depend on secondary data from past events.

For data intended for trend analysis, studies have recommended the application of nonparametric techniques such as the Seasonal Kendal test where transformation techniques do not yield symmetric data [ 19 ]. Should a parametric test be preferred on a data-set that includes outliers, practitioners may evaluate the influence of outliers by performing the test twice, once using the full data-set (including the outliers) and again on the reduced data-set (excluding the outliers). If the conclusions are essentially the same, then the suspect datum may be retained, failing which a nonparametric test is recommended.

4. Missing values

While most statistical methods presumes a complete data-set for analysis, missing values are frequently encountered problems in water quality studies [ 35 , 36 ]. Handling missing values can be a challenge as it requires a careful examination of the data to identify the type and pattern of missingness, and also have a clear understanding of the most appropriate imputation method. Gaps in water quality data-sets may arise due to several reasons, among which are imperfect data entry, equipment error, loss of sample before analysis and incorrect measurements [ 37 ]. Missing values complicate data analysis, cause loss of statistical efficiency and reduces statistical estimation power [ 37 - 39 ]. For data intended for time-series analysis and model building, gaps become a significant obstacle since both generally require continuous data [ 40 , 41 ]. Any estimation of missing values should be done in a manner that minimise the introduction of more bias in order to preserve the structure of original data-set [ 41 , 42 ].

The best way to estimate missing values is to repeat the experiment and produce a complete data-set. This option is however, not feasible when conducting a retrospective study since it depend on historical data. Where it is not possible to re-sample, a model or non-model techniques may be applied to estimate missing values [ 43 ].

If the proportion of missing values is relatively small, listwise deletion has been recommended. This approach, which is considered the easiest and simplest, discards the entire case where any of the variables are missing. Its major advantage is that it produce a complete data-set, which in turn allows for the use of standard analysis techniques [ 44 ]. The method also does not require special computational techniques. However, as the proportion of missing data increases, deletion tends to introduce biasness and inaccuracies in subsequent analyses. This tends to reduce the power of significance test and is more pronounced particularly if the pattern of missing data is not completely random. Furthermore, listwise deletion also decreases the sample size which tends to reduce the ability to detect a true association. For example, suppose a data-set with 1,000 samples and 20 variables and each of the variables has missing data for 5% of the cases, then, one could expect to have complete data for only about 360 individuals, thus discarding the other 640.

On the other hand, pairwise deletion removes incomplete cases on an analysis-by-analysis basis, such that any given case may contribute to some analyses but not to others [ 44 ]. This approach is considered an improvement over listwise deletion because it minimises the number of cases discarded in any given analysis. However, it also tend to produce bias if the data is not completely random.

Several studies have applied imputation techniques to estimate missing values. A common assumption with these methods is that data should be missing randomly [ 45 ]. The most common and easiest imputation technique is replacing the missing values with an arithmetic mean for the rest of the data [ 35 , 41 ]. This is recommended where the frequency distribution of a variable is reasonably symmetric, or has been made so by data transformation methods. The advantage of arithmetic mean imputation is generation of unbiased estimates if the data is completely random because the mean lands on the regression line. Even though the insertion of mean value does not add information, it tends to improve subsequent analysis. However, while simple to execute, this method does not take into consideration the subjects patterns of scores across all the other variables. It changes the distribution of the original data by narrowing the variance [ 46 ]. If the data assumes an asymmetric distribution, the median has been recommended as a more representative estimate of the central tendency and should be used instead of the mean.

[ 47 ], recommended model-based substitution techniques as more flexible and less ad hoc approach of estimating missing values as compared to non-model methods. A simple modelling technique is to regress the previous observations into an equation which estimates missing values [ 35 , 48 ]. The time-series auto-regressive model has been described as an improvement and more accurate method of estimating missing values [ 25 ]. Unlike the arithmetic mean and median replacement methods, regression imputation techniques estimates missing values of a given variable using data of other parameters. This tends to reduce the variance problem, which is common with the arithmetic mean imputation and median replacement methods [ 41 , 49 ].

On the other hand, the maximum likelihood technique uses all the available complete and incomplete data to identify the parameter values that have the highest probability of producing the sample data [ 44 ]. It runs a series of data iterations by replacing different values for the unknown parameters and converges to a single set of parameters with the highest probability of matching the observed data [ 41 ]. The method has been recommended as it tends to give efficient estimates with correct standard errors. However, just like other imputation methods, the maximum likelihood estimates can be heavily biased if the sample size is small. In addition, the technique requires a specialised software which may be expensive, challenging to use and time consuming.

Some studies have considered the relationship between parameters as an effective approach of estimating missing values [ 50 ]. For instance, missing conductivity values can be calculated from the total dissolved solids value (TDS) by a simple linear regression where p-value and r-value are known to exist and the missing value lies between the two variables. Equation 2 , where a is in the range 1.2-1.8, has been described as an equally important estimator of missing conductivity values [ 1 , 51 ].

The constant, a , is high in water of high chloride and low sulphate [ 51 ]. [ 52 ] estimated missing potassium values by using a linear relationship between potassium and sodium. The relationship gave a high correlation coefficient of 0.904 (p<0.001).

As of late, research has explored the application of artificial intelligence (AI) techniques to handle missing values in the water quality sector. Among the major AI techniques that have been applied is the Artificial Neural Networks (ANN) and Hybrid Evolutionary Algorithms (HEA) (48, 53, 54). Nevertheless, it should also be highlighted that all techniques for estimating missing values invariably affect the results. This is more pronounced when missing values characterise a significant proportion of the data being analysed. A research should thus consider the sample size when choosing the most appropriate imputation method.

5. Scientific facts

The integrity of water quality data can also be assessed by checking whether the results are inline with known scientific facts. To ascertain that, a researcher must have some scientific knowledge regarding the characteristics of water quality variables. Below are some scientific facts that can be used to assess data integrity [ 1 ].

Presence of nitrate in the absence of dissolved oxygen may indicate an error since nitrate is rapidly reduced in the absence of oxygen. The dissolved oxygen meter might have malfunctioned or oxygen might have escaped from the sample before analysis.

Component parts of a water-quality variable must not be greater than the total variable. For example:

Total phosphorus ≥Total dissolved phosphorus>Ortho-phosphate.

Total Kjeldahl nitrogen ≥Total dissolved Kjeldahl nitrogen>ammonia.

Total organic carbon ≥Dissolved organic carbon.

Species in a water body should be described correctly with regards to original pH of the water sample. For example, carbonate species will normally exist as HCO 3 - while CO 3 2- cannot co-exist with H 2 CO 3.

6. Censored data

A common problem faced by researchers analysing environmental data is the presence of observations reported to have non-detectable levels of a contaminant. Data which are either less than the lower detection limit, or greater than the upper detection limit of the analytical method applied are normally artificially curtailed at the end of a distribution, and are termed “censored values” [ 14 ]. Multiple censored results may be recorded when the laboratory has changed levels of detection, possibly as a result of an instrument having gained more accuracy, or the laboratory protocol having established new limits. If the values are below the detection limit, they are abbreviated as BDL, and when above the limit, as ADL [ 55 , 56 ].

Various methods of treating censored values have been developed to reduce the complication generally brought about by censored values [ 57 ]. The application of an incorrect method may introduce bias especially when estimating the mean and variance of data distribution [ 58 ]. This may consequently distort the regression coefficients and their standard errors, and further reduce the hypothesis testing power. A researcher must thus decide on the most appropriate method to analyse censored values. One might reason that since these values are extra ordinarly small, they are not important and discard them while some might be tempted to remove them inorder to ease statistical analysis. Deletion has however been described as the worst practise as it tends to introduce a strong upward bias of the central tendency which lead to inaccurate interpretation of data [ 19 , 59 - 62 ].

The relatively easiest and most common method of handling censored values is to replace them with a real number value so that they conform to the rest of data. The United State Environmental Agency suggested substitution if censored data is less than 15% of the total data-set (63, 64). [ 8 ], BDL, for example x < 1.1, were multiplied by the factor 0.75 to give 0.825. ADL values, for example 500 < x, were recorded as one magnitude higher than the limit values to give 501. [ 65 ] recommended substituting with 1   2 DL or 1 2 DL if the sample size is less than 20 and contains less than 45% of its data as censored values. [ 66 ] suggested substitution by 1 √ 2 D L if the data are not highly skewed and substitution by 1 2 DL otherwise. [ 67 ], however, criticised the substitution approach and illustrated how the practice could produces poor estimates of correlation coefficients and regression slopes. [ 68 ] further explained that substitution is not suitable if the data has multiple detection limits [ 68 , 69 ].

A second approach for handling censored values is the maximum likelihood estimation (MLE). It is recommended for a large data-set which assumes normality and contains censored results [ 38 , 65 , 70 , 71 ]. This approach basically uses the statistical properties of non-censored portion of the data-set, and an iterative process to determine the means and variance. The MLE technique generates an equation that calculates mean and standard deviation from values assumed to represent both the detects and non-detect results [ 69 ]. The equation can be used to estimate values that can replace censored values. However, the technique is reportedly ineffective for a small data-set that has fewer than 50 BDLs [ 69 ].

When data assumes an independent distribution and contain censored values, non-parametric methods like the Kaplan-Meir method, can be considered for analysis [ 59 ]. The Kaplan-Meir method creates an estimate of the population mean and standard deviation, which is adjusted for data censoring, based on the fitted distribution model. Just like any non-parametric techniques for analysing censored data, the Kaplan-Meier is only applicable to right-censored results (i.e. greater than) [ 72 ]. To use Kaplan-Meier on left-censored values, the censored values must be converted to right-censored by flipping them over to the largest observed value [ 65 , 71 , 72 ]. To ease the process, [ 73 ] have developed a computer program that does the conversion. [ 71 ], however, found the Kaplan-Meier method to be effective when summarising a data-set containing up to 70% of censored results.

In between the parametric and non-parametric methods is a robust technique called Regression on Order Statistics (ROS) [ 38 ]. It treats BDLs based on the probability plot of detects. The technique is applicable where the response variable (concentration) is a linear function of the explanatory variable (the normal quartiles) and if the error variance of the model is constant. It also assumes that all censoring thresholds are left-censored and is effective for a data-set which contains up to 80% censored values [ 59 ]. The ROS technique uses data plots on a modelling distribution to predict censored values. [ 59 ] and [ 68 ] evaluated ROS as a reliable method for summarising multiply censored data. Helsel and Cohn (38) also described ROS as a better estimator of the mean and standard deviation as compared to MLE, when the sample size is less than 50 and contains censored values.

7. Statistical methods

The success of an analysis of water quality data primarily depends on the selection of the right statistical method which considers common data characteristics such as normality, seasonality, outliers, missing values, censoring, etc., [ 74 ]. If the data assumes an understandable and describable distribution, parametric methods can be used [ 14 ]. However, non-parametric techniques are slowly replacing parametric techniques mainly because the latter are sensitive to common water characteristics like outliers, missing values and censored value [ 75 ].

7.1. Computer application in data treatment

The increase in various computer programs has made it easy to detect and treat erroneous data. Computers now provide flexibility and speedy methods of data analysis, tabulation, graph preparation or running models, among others. Various software such as Microsoft Excel, Minitab, Stata and MATLAB have become indispensable tools for analysing environmental data. These software perform various computations associated with checking assumptions about statistical distributions, error detection and their treatment. However, the major problem encountered by researchers, is lack of guidance regarding selection of the most appropriate software. Computer-aided statistical analysis should be undertaken with some understanding of the techniques being used. For example, some statistical software packages might replace missing values with the means of the variable, or prompt the user for case-wise deletion of analytical data, both of which might be considered undesirable [ 52 ].

Lately, machine learning algorithms like the artificial neural networks (ANNs) [ 67 , 76 - 78 ], and genetic algorithms (GA) [ 76 , 79 ] have gained momentum in water quality monitoring studies. [ 41 ] pointed out that these technique generally yields the best parameter estimates in the data set with the least amount of missing data. Nevertheless, as the percentage of missing data increases, the performance of ANN which is generally measured by the errors in the parameter estimates, decreases and may reach performance levels similar to those obtained by the general substitution methods. However, in all cases the effectiveness of these methods lies on the user’s ability to manipulate and display data correctly.

8. Conclusion

This chapter discussed the common data characteristics which tend to affect statistical analysis. It is recommended that practitioners should explore for outliers, missing values and censored values in a data-set before undertaking in-depth analysis. Although an analyst might not be able to establish the causal of such characteristic, eliminate or overcome some of the errors, having knowledge of their existence assists in establishing some level of confidence in drawing meaningful conclusions. It is recommended that water quality monitoring programs should strive to collect data of high quality. Common methods of ascertaining data quality are practising duplicate samples, using blanks or reference samples, and running performance audits. If a researcher is not sure of how to treat a characteristic of interest, a non-parametric method like Seasonal Kendal test could provide a better alternative since it is insensitive to common water quality data characteristics like outliers.

• 1. Steel A, Clarke M, Whitfield P. Use and Reporting of Monitoring Data. In: Bartram J, Ballance R, editors. Water Quality Monitoring-A Practical Guide to the Design and Implementation of Freshwater Quality Studies and Monitoring Programmes: United Nations Environment Programme and the World Health Organization; 1996.
• 2. Mitchell P. Guidelines for Quality Assurance and Quality Control in Surface Water Quality Programs in Alberta. Alberta Environment, 2006.
• 3. Tasić S, Feruh MB. Errors and Issues in Secondary Data used in marketing research. The Scientific Journal for Theory and Practice of Socioeconomic Development. 2012;1(2).
• 4. Doong DJ, Chen SH, Kao CC, Lee BC, Yeh SP. Data quality check procedures of an operational coastal ocean monitoring network. Ocean Engineering. 2007;34(2):234-46.
• 5. Taylor S, Bogdan R. Introduction to research methods: New York: Wiley; 1984.
• 6. Wahi MM, Parks DV, Skeate RC, Goldin SB. Reducing errors from the electronic transcription of data collected on paper forms: a research data case study. Journal of the American Medical Informatics Association. 2008;15(3):386-9.
• 7. UNEP-WHO. Use and Reporting of monitoring data. In: Bartram J, Ballace R, editors. Water Quality Monitoring A practical Guide to the Design and Implementation of Freshwater Quality Studies and Monitoring Programmes. London: United Nations Environmental program, World Health Organisation; 1996.
• 8. Dzwairo B. Modelling raw water quality variability in order to predict cost of water treatment. Pretoria: Tshwane University of Technology; 2011.
• 9. Kawado M, Hinotsu S, Matsuyama Y, Yamaguchi T, Hashimoto S, Ohashi Y. A comparison of error detection rates between the reading aloud method and the double data entry method. Controlled Clinical Trials. 24. 2003:560–9.
• 10. Cummings J, Masten J. Customized dual data entry for computerized analysis. Quality Assurance: Good Practice, Regulation, and Law. 1994;3:300-3.
• 11. Brown ML, Austen DJ. Data management and statistical techniques. Murphy BR, Willis DW, editors. Bethesda, Maryland: America Fisheries Society; 1996.
• 12. Rajaraman V. Self study guide to Analysis and Design of Information Systems. New Delhi: Asoke K. Ghosh; 2006.
• 13. Data Analysis and Interpretation. Rhe Monitoring Guideline. Austria2000.
• 14. Helsel DR, Hirsch RM. Statistical Methods in Water Resources. Amsterdam, Netherlands: Elsevier Science Publisher B.V; 1992.
• 15. Köster D, Hutchinson N. Review of Long-Term Water Quality Data for the Lake System Health Program. Ontario: 2008 Contract No.: GLL 80398.
• 16. Iglewicz B, Hoaglin DC. How to Detect and Handle Outliers1993.
• 17. Zar JH. Biostatistical Analysis. Upper Saddle River, NJ.: Prentice-Hall Inc; 1996.
• 18. Silva-Ramírez E-L, Pino-Mejías R, López-Coello M, Cubiles-de-la-Vega M-D. Missing value imputation on missing completely at random data using multilayer perceptrons. Neural Networks. 2011;24(1):121-9.
• 19. USEPA., Ecology. Do. Technical Guidance for Exploring TMDL Effectiveness Monitoring Data. 2011.
• 20. Stoimenova E, Mateev P, Dobreva M. Outlier detection as a method for knowledge extraction from digital resources. Review of the National Center for Digitization. 2006;9:1-11.
• 21. US-EPA. Statistical Analysis of Groundwater Monitoring Data at RCRA Facilities, Unified Guidance (Unified Guidance). US-Environmetal Protection Agency, 2009.
• 22. US-EPA. Data Quality Assessment: Statistical Methods for Practitioners. In: Agency USEP, editor: United States Environmental Protection Agency; 2006.
• 23. Osoborne JW. Improving your data transformations: Applying the Box-Cox transformation. Practical Assessment, Research and Evaluation. 2010;15(12).
• 24. High R. Dealing with 'Outlier': How to Maintain your data's integrity.
• 25. Chi Fung DS. Methods for the Estimation of Missing Values in Time Series. Western Australia: Edith Cowan University; 2006.
• 26. US-EPA. Statistical Training Course for Ground-Water Monitoring Data Analysis.. Washington D.C.: Environmental Protection Agency, 1992.
• 27. Grubbs FE, Beck G. Extension of sample sizes and percentage points for significance tests of outlying observations. Technometric. 1972;14:847-54.
• 28. Tiwari RC, Dienes TP. The Kalman filter model and Bayesian outlier detection for time series analysis of BOD data. Ecological Modelling 1994;73:159-65.
• 29. De Muth JE. Basic statistics and pharmaceutical statistical applications: CRC Press; 2014.
• 30. Gibbons RD. Statistical Methods for Groundwater Monitoring. New York: John Wiley & Sons; 1994.
• 31. Walfish S. A Review of Statistical Outlier Methods. Pharmaceutical Technology. 2006.
• 32. Robinson RB, Cox CD, Odom K. Identifying Outliers in Correlated Water Quality Data. Journal of Environmental Engineering. 2005;131(4):651-7.
• 33. Filzmoser P, editor A multivariate outlier detection method. Proceedings of the seventh international conference on computer data analysis and modeling; 2004: Minsk: Belarusian State University.
• 34. Garret D. The Chi-square plot: A tool for multivariate recognition. Journal of Geochemical Exploration. 1989;32:319-41.
• 35. Ssali G, Marwala T. Estimation of Missing Data Using Computational Intelligence and Decision Trees. n.d.
• 36. Noor NM, Zainudin ML. A Review: Missing Values in Environmental Data Sets. International Conference on Environment 2008 (ICENV 2008); Pulau Pinang2008.
• 37. Calcagno G, Staiano A, Fortunato G, Brescia-Morra V, Salvatore E, Liguori R, et al. A multilayer perceptron neural network-based approach for the identification of responsiveness to interferon therapy in multiple sclerosis patients. Information Sciences. 2010;180(21):4153-63.
• 38. Helsel DR, Cohn T. Estimation of description statistics for multiply censored water quality data. Water Resource Research. 1988; 24:1997-2004.
• 39. Little RJ, Rubin DB. Statistical analysis with missing data. New York: John Wiley and Sons; 1987.
• 40. Junninen H, Niska H, Tuppurainen K, Ruuskanen J, Kolehmainen M. Methods for imputation of missing values in air quality data sets. Atmospheric Environment. 2004; 38:2895–907.
• 41. Nieh C. Using Mass Balance, Factor Analysis, and Multiple Imputation to Assess Health Effects of Water Quality. Chicago, Illinois: University of Illinois 2011.
• 42. Luengo J, Garcia S, Herrera F. A study on the use of imputation methods for experimentation with Radial Basis Function Network classifiers handling missing attribute values: The good synergy between RBFNs and EventCovering methodI. Neural Networks. 2010; 23:406-18.
• 43. Lakshminarayan K, Harp S, Samad T. Imputation of missing data in industrial databases. Applied Intelligence. 1999;11(3):259-75.
• 44. Baraldi AN, Enders CK. An introduction to modern missing data analyses. Journal of School Psychology. 2010;48(1):5-37.
• 45. Rubin DB. Inference and missing data. Biometrika. 1976;63(3):581-92.
• 46. Enders CK. A primer on the use of modern missing-data methods in psychosomatic medicine research. PsychosomaticMedicine. 2006;68:427–36.
• 47. Fogarty DJ. Multiple imputation as a missing data approach to reject inference on consumer credit scoring. Intersat. 2006.
• 48. Smits A, Baggelaar PK. Estimating missing values in time series. Netherlands: Association of River Waterworks – RIWA; 2010.
• 49. Sartori N, Salvan A, Thomaseth K. Multiple imputation of missing values in a cancer mortality analysis with estimated exposure dose. ComputationalStatistics and DataAnalysis. 2005;49(3):937–53.
• 50. Güler C, Thyne GD, McCray JE, Turner AK. Evaluation of graphical and multivariate statistical methods for classification of water chemistry data. Hydrogeology 2002;10:455–74.
• 51. Dzwairo B, Otieno FAO, Ochieng' GM. Incorporating surface raw water quality into the cost chain for water services: Vaal catchment, South Africa. Research Journal of Chemistry and Environment. 2010;14(1):29-35.
• 52. Guler C, Thyne GD, McCray JE, Turner AK. Evaluation of graphical and multivariate statistical methods for classification of water chemistry data. Hydrogeology. 2002;10:455-74.
• 53. Starret KS, Starret KS, Heier T, Su Y, Tuan D, Bandurraga M. Filling in Missing Peakflow data using Artificial Neural Networks. Journal of Engineering and Applied Science. 2010;5(1).
• 54. Aitkenhead MJ, Coull MC. An application based on neural networks for replacing missing data in large datasets n.d.
• 55. Lin P-E, Niu X-F. Comparison of Statistical Methods In Handling Minimum Detection Limits. Department of Statistics,, Florida State University, 1998.
• 56. Darken PF. Testing for changes in trend in water quality data. Blacksburg, Virginia: Virginia Polytechnic Institute and State University; 1999.
• 57. Farnham IM, Stetzenbach KJ, Singh AS, Johannesson KH. Treatment of nondetects in multivariate analysis of groundwatervgeochemistry data. Chemometrics Intelligent Lab Sys. 2002;60(265-281).
• 58. Lyles RH, Fan D, Chuachoowong R. Correlation coefficient estimation involving a left censored laboratory assay variable. Statist Med. 2001;20:2921-33.
• 59. Lopaka L, Helsel D. Statistical analysis of water-quality data containing multiple detection limits: S-language software for regression on order statistics. Computers & Geosciences. 2005;31:1241-8.
• 60. Gheyas IA, Smith LS. A neural network-based framework for the reconstruction of incomplete data sets. Neurocomputing. 2010;73(16–18):3039-65.
• 61. Nishanth KJ, Ravi V, Ankaiah N, Bose I. Soft computing based imputation and hybrid data and text mining: The case of predicting the severity of phishing alerts. Expert Systems with Applications. 2012;39(12):10583-9.
• 62. Helsel D. Fabricating data: How substituting values for nondetects can ruin results, and what can be done about it. Chemosphere. 2006;65:2434-9.
• 63. Kalderstam J, Edén P, Bendahl P-O, Strand C, Fernö M, Ohlsson M. Training artificial neural networks directly on the concordance index for censored data using genetic algorithms. Artificial Intelligence in Medicine. 2013;58(2):125-32.
• 64. Environmental Protection Agency. Data Quality Assessment: Statistical Methods for Practitioners. Agency USEP, editor. Washington: United States Environmental Protection Agency, Office of Environmental Information, 2006; 2006 14 June 2013. 198 p.
• 65. Hewett P, Ganser GH. A Comparison of Several Methods for Analyzing Censored Data. British Occupational Hygiene. 2007;57(7):611-32.
• 66. Hornung RW, Reed LD. Estimation of Average Concentration in the Presence of Nondetectable Values. Applied Occupational and Environmental Hygiene. 1990;5(1):46-51.
• 67. Kang P. Locally linear reconstruction based missing value imputation for supervised learning. Neurocomputing. 2013(0).
• 68. Shumway RH, Azari RS, Kayhanian M. Statistical approaches to estimating mean water quality concentrations with detection limits. Environmental Science and Technology. 2002;36:3345-53.
• 69. Helsel D. Less than obvious-statistical treatment data below the detection limit. Environmental Science and Technology. 1990;24(12):1766-44.
• 70. Sanford RF, Pierson CT, Crovelli RA. An objective replacement method for censored geochemical data. Math Geol. 1993;25:59-80.
• 71. Antweiler RC, Taylor HE. Evaluation of Statistical Treatments of Left-Censored Environmental Data using Coincident Uncensored Data Sets. Environmental Science and Technology. 2008;42(10):3732-8.
• 72. Fu L, Wang Y-G. Statistical tools for analysing water quality data2012.
• 73. Silva JdA, Hruschka ER. An experimental study on the use of nearest neighbor-based imputation algorithms for classification tasks. Data & Knowledge Engineering. 2013;84(0):47-58.
• 74. Visser A, Dubus I, Broers HP, Brouyere S, Korcz M, Orban P, et al. Comparison of methods for the detection and extrapolation of trends in groundwater quality. Journal of Environmental Monitoring. 2009;11(11):2030-43.
• 75. Schertzer TL, Alexander RB, Ohe DJ. The Computer Program Estimate Trend (ESTREND), a system for the detection of trends in water-quality data Water Resources investigation report. 1991;91-4040:56-7.
• 76. Recknagel F, Bobbin J, Whigham P, Wilson H. Comparative application of artificial neural networks and genetic algorithms for multivariate time-series modelling of algal blooms in freshwater lakes. Journal of Hydroinformatics. 2002;4(2):125-34.
• 77. Lee JHW, Huang Y, Dickman M, Jayawardena AW. Neural network modelling of coastal algal blooms.Ecological Modelling. 2003;159(179-201).
• 78. Singh KP, Basant A, Malik A, Jain G. Artificial neural network modeling of the river water quality—a case study. Ecological Modelling. 2009;220(6):888-95.
• 79. Muttil N, Lee JHW. Genetic programming for analysis and real-time prediction of coastal algal blooms. Ecological Modelling. 2005;189:363-76.

© 2015 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution 3.0 License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Continue reading from the same book

Published: 09 September 2015

By Nataša Gros

1811 downloads

By Marco André da Silva Costa and Magda Sofia Valério...

2191 downloads

By A.M. Taiwo, A.T. Towolawi, A.A. Olanigan, O.O. Olu...

3010 downloads

IntechOpen Author/Editor? To get your discount, log in .

Discounts available on purchase of multiple copies. View rates

Local taxes (VAT) are calculated in later steps, if applicable.

Support: [email protected]

• Geospatial Applications
• Environment & Climate

Mapping of Water Quality Index (WQI) using Geographical Information System (GIS) as Decision Supporting System Tool

K. Veerabhadram Assistant Professor, Department of Environmental Studies, College of Engineering GITAM, Visakhapatnam-530 045, India

Contamination of drinking water has become a major challenge to the Environmentalist in the rapid developing countries. As more and more people are exposed to contamination of drinking water, many issues arise that not only involve premeditating the contaminated water, but also preventing similar situations from occurring future. Te drinking water is contaminate through the pipe distribution system or directly through ground water. By mapping using the Decision Support system like GIS can be useful for taking quick decision as graphical representation would be easy to take decision by the policy makers.

A water quality index provides a single number (like a grade) that expresses overall water quality at a certain location and time based on several water quality parameters. The objective of an index is to turn complex water quality data into information that is understandable and useable by the public. A single number cannot tell the whole story of water quality; there are many other water quality parameters that are not included in the index. The WQI presented is not specifically aimed at human health or aquatic life regulations.

Assistant Professor, Department of Environmental Studies, College of Engineering, GITAM, Visakhapatnam-530 045, India

Decision makers in environmental fields face the difficult challenges of anticipating the potential biophysical and socioeconomic impacts of managements and policy interventions over regions that may vary dramatically in terms of climates, soils, topography, land use and other factors. Leung(1997) addressed a host of conceptual, theoretical, systems.

However, a water index based on some very important parameters can provide a simple indicator of water quality. It gives the public a general idea the possible problems with the water in the region.

Objective of Water Quality Index The main objective of Water Quality Index is to turn complex water quality data into information that is understandable and useable by the public. Water Quality Index based on some very important parameters can provide a simple indicator of water quality. it gives the public a general idea of the possible problems with water in a particular region. The indices are among the most effective ways to communicate the information on water quality trends to the public or to the policy makers and water quality management. It is also defined as a rating reflecting the composite influence of different water quality parameters on the overall quality of water .the concept of indices to represent gradation in water quality

WQI Calculation For calculation of WQI, section of parameters has great importance. Since selection of too many parameters might widen the quality index and importance of various parameters depends on the intended use of water, twelve physico-chemical parameters namely pH, Conductivity, Turbidity, Total hardness, Mg hardness, Ca hardness, TDS, Chlorides, Alkalinity, Sulphate, Nitrates, Iron, were used to calculate WQI.

Importance of GIS as Decision Support System A spatial Decision Support System ( SDSS) is a computer -based system designed to assist decision system. Typically, such a system will include spatial data relevant to the decision, analytic tools to process the data in ways meaningful for decision makers, and output or display functions. Thus, an SDSS has considerable overlap with the functionality of a Geographical Information System (GIS). According to the National Center of Geographic Information and Analysis (NCGIA) an SDSS is an ” interactive, computer-based system designed to support a user or group of users is achieving a higher effectiveness of decision making while solving a semi-structured spatial decision problem”.

Since there are many examples of SDSS, for specific decisions in the environmental domain, particularly in the areas of water pollution, crop, livestocks, flood, and forest management. National reliance of Ground Water has increased exponentially over the past three decades. Although contamination of Ground Water has been occurring for many centuries, the current concern for control of water pollution and for maintenance of high quality waster supplies has stimulated an interest in the protection of Ground Water resources from contamination. The level of Ground Water contamination depends on the physical characteristics of the area, the chemical nature of the pollutant and the method of application. This paper gives a case study of an area appraises of a domain of the study area, which incorporates the influence of the physical characteristics of any area of evaluate an index that reflects the vulnerability of an area towards of Ground Water pollution.

Objectives of the proposed project The main objectives of the project work are as follows.

To develop an expert system for decision makers and prepared a software by which large number of data can be handled in a user friendly manner

RELATED ARTICLES MORE FROM AUTHOR

The making of Jack Dangermond: Godfather of GIS

Why You Need Location Intelligence for 5G Deployment Success

UMass Global launches Institute for Geospatial Education

© Geospatial Media and Communications. All Rights Reserved.​

Assessment of groundwater quality

• October 2017
• Publisher: e-PG Pathshala, UGC, MHRD, Govt. of India

• University of Delhi

Abstract and Figures

Discover the world's research

• 25+ million members
• 160+ million publication pages
• 2.3+ billion citations

• Jagadish Ku Tripathy

• L. A. RICHARDS
• D W Westcot
• Larry W Mays
• Recruit researchers
• Join for free
• Login Email Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google Welcome back! Please log in. Email · Hint Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google No account? Sign up

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: 05 September 2024

Remote sensing analysis of spatiotemporal impacts of anthropogenic influence on mountain landscape ecology in Pir Chinasi national park

• Muhammad Akhlaq Farooq 1 ,
• Muhammad Asad Ghufran 1 ,
• Naeem Ahmed 2 ,
• Kotb A. Attia 3 ,
• Arif Ahmed Mohammed 3 ,
• Yaser M. Hafeez 4 ,
• Aamir Amanat 2 ,
• Muhammad Shahbaz Farooq 5 ,
• Muhammad Uzair 6 &
• Saima Naz 7  

Scientific Reports volume  14 , Article number:  20695 ( 2024 ) Cite this article

Metrics details

• Climate sciences
• Environmental sciences

Mountain landscapes can be fragmented due to various human activities such as tourism, road construction, urbanization, and agriculture. It can also be due to natural factors such as flash floods, glacial lake outbursts, land sliding, and climate change such as rising temperatures, heavy rains, or drought.The study’s objective was to analyze the mountain landscape ecology of Pir Chinasi National Park under anthropogenic influence and investigate the impact of anthropogenic activities on the vegetation. This study observed spatiotemporal changes in vegetation due to human activities and associated climate change for the past 25 years (1995–2020) around Pir Chinasi National Park, Muzaffrabad, Pakistan. A structured questionnaire was distributed to 200 residents to evaluate their perceptions of land use and its effects on local vegetation. The findings reveal that 60% of respondents perceived spatiotemporal pressure on the park. On the other hand, the Landsat-oriented Normalized Difference Vegetation Index (NDVI) was utilized for the less than 10% cloud-covered images of Landsat 5, 7, and 8 to investigate the vegetation degradation trends of the study area. During the entire study period, the mean maximum NDVI was approximately 0.28 in 1995, whereas the mean minimum NDVI was − 2.8 in 2010. QGIS 3.8.2 was used for the data presentation. The impact of temperature on vegetation was also investigated for the study period and increasing temperature trends were observed. The study found that 10.81% (1469.08 km 2 ) of the area experienced substantial deterioration, while 23.57% (3202.39 km 2 ) experienced minor degradation. The total area of degraded lands was 34.38% (or 4671.47 km 2 ). A marginal improvement in plant cover was observed in 24.88% of the regions, while 9.69% of the regions experienced a major improvement. According to the NDVI-Rainfall relationships, the area was found to be significantly impacted by human pressures and activities (r ≤ 0.50) driving vegetation changes covering 24.67% of the total area (3352.03 km 2 ). The area under the influence of climatic variability and change (r ≥ 0.50 ≥ 0.90) accounted for 55.84% (7587.26 km 2 ), and the area under both climatic and human stressors (r ≥ 0.50 < 0.70) was 64%. Sustainable land management practices of conservation tillage, integrated pest management, and agroforestry help preserve soil health, water quality, and biodiversity while reducing erosion, pollution, and the degradation of natural resources. landscape restoration projects of reforestation, wetland restoration, soil erosion control, and the removal of invasive species are essential to achieve land degradation neutrality at the watershed scale.

Similar content being viewed by others

Land use and land cover changes influence the land surface temperature and vegetation in Penang Island, Peninsular Malaysia

Impacts of spatio-temporal change of landscape patterns on habitat quality across Zayanderud Dam watershed in central Iran

Variations in urban land surface temperature intensity over four cities in different ecological zones

Introduction.

Vegetation is an essential component of the biological cycle. Vegetation trends highlight changes in ecological systems and help determine the best strategies for mitigating climate change. Highlighting the changes in vegetation due to anthropogenic activities as well as climate change is essential for improvement. An important topic in the study of the local ecological environment is the components of the ecosystem represented by vegetation under the effect of anthropogenic activities 1 . The dynamics of regional vegetation have a substantial impact on ecological security 2 , ecosystem services and, are frequently used as important indicators of ecological variations in the environment 3 . As human activity has risen in recent decades, alterations in vegetation has deeply captured the trails of human activity, which have been made worse by climate change 4 , 5 . Anthropogenic stresses are thought to have significant effects on vegetation and ecosystem services as we go through the Anthropocene 6 , 7 . The global ecosystem is now changing as a result of climate change and land cover variations, which are identified as key influences on the dynamics of vegetation under universal change. Numerous anthropogenic stresses along with climate change are causing vegetation to alter and degrade in highland locations, which further compromises the ecosystem services provided by mountains and the way of life of a small number of mountain people. In the realm of research on global change, understanding the linkages between natural vegetation and cultivated vegetation has been a crucial problem that is receiving more and more attention from the scientific community 8 . To investigate the change in vegetation, Normalized Difference Vegetation Index (NDVI) is mostly recommended 1 , 9 , 10 . Utilizing the NDVI, a surrogate technique for the greenness of landscape and biological dynamics of change has made satellite Remote Sensing (RS) of vegetation simple 1 , 9 , 10 . The investigation of vegetation status under changing climatic patterns and the tracking of the ecological environment’s quality relies heavily on the monitoring and assessment of NDVI changes in vegetation 11 . Limited research examined how vegetation dynamics relate to both human activity and climate change 1 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 . Moreover, residents of these mountains can better highlight the changes in vegetation that occurred due to human intervention as well as climate change 22 , 23 , 24 . The objective of this study was to understand the mechanisms behind grassland degradation of Pir Chinasi National Park and managing damaged grasslands, which needs quantitative assessments of the relative effects of anthropogenic activity and climate change on grasslands 25 . As for the significance of the mountain, the UN 2030 Sustainable Development Agenda has included targets for protecting and sustainably developing mountain ecosystems 26 . The cover index (SDG 15.4.2) of the mountain forests of Pir Chinasi National Park has been utilized as a first approximation for measuring the sustainability and protection of mountain habitats as well as the progress made towards the aforementioned aim.

As the heights vary from roughly 700 m Above Sea Level (ASL), the Pir Chinasi National Park in the Muzaffarabad (Fig.  1 ) neighborhood exhibits a topographic variety 27 . The study area’s height is 2900 m (9,500 ft) 28 . The area has natural beauty with its dense trees, swiftly flowing rivers, and twisting streams. The mean extreme temperature for January and June respectively, ranged from − 2.6 °C to 45.2 °C. The average rainfall is found to be between 1000 and 1300 mm or around 680 mm and falls over the four months from May to August. During the day, the wind blows from west to east, while at night, it blows from southeast to north and the wind is brisker in the afternoon. The area of Pir Chinasi is found to be dry Subtropical type in which Acacia modesta, Olea ferruginea, and Chir Pine are the most dominant tree species in the area 29 . The vegetation of the area is comprised of a spacious variety of herbs, trees, climbers, and shrubs. The ground cover is comprised of a wide variety of angiosperms, mosses, and ferns.

Location map of the Pir Chinasi National Park, Muzaffarabad (District of Azad Jammu & Kashmir). The map was created using ArcGIS Desktop software (version 10.8.2, Esri lnc., Redlands,CA, USA).

The western Himalayan region, a biodiversity hotspot, is impacted in several ways by anthropogenic climate change. Due to the great seasonality of the western Himalayan region, including the Muzaffarabad district, investigations on the interactions between the timing of phenological periodic occurrences and climatic seasonality are important. The different land users of Pir Chinasi National Park benefit from the area’s natural resources and environment. During dry seasons, herders who live on the mountain frequently light fires to encourage the growth of forage for cattle grazing. This study assesses the dynamics of vegetation degradation during five years (1995–2020). This serves as the baseline information for tracking the watershed-level progress towards the objectives and pathways of SDGs 15.3.1 and 15.4.2 related to land degradation neutrality.

Materials and methods

Data collection.

This study was conducted in Pir Chinasi National Park, Muzaffarabad, Kashmir, Pakistan. Pir Chinasi National Park is a tourist spot, and many tourists visit this area annually. A one-shot Household Survey (HHS) was conducted from 200 residents using convenient sampling in this study. It was a hilly area with a scattered population, so it was difficult to collect a larger sample size. A team of enumerators and the chief investigator spoke with the heads of the families. Family members were contacted in case of the unavailability of the head of household. The purpose of each question of the survey was explained to the respondents. Participants were provided with informed consent, and they were informed of the study’s advantages, goals, and financing. All surveys were conducted following the shared research principles and ethics. The survey asked about households’ occupation, age, gender, education, environmental harm to vegetation, and government support. Moreover, plants/seeds were handled under the direct supervision of Dr. Naeem Ahmed, National University of Modern Languages, Islamabad-Pakistan following the proper national and international strategies.

Data processing

The authors confirmed that the collection and execution of the experiment complied with the IUCN statement on research, involving species at risk of extinction and by the convention on the trade in endangered species of wild fauna and flora. For data collection, Google Forms were utilized and for data processing and analysis, SPSS was used. All methods were performed following the relevant guidelines and regulations.

To investigate the effects of human activities on the flora and ecology of the area, Remote Sensing (RS) and Geographic Information System (GIS) techniques and datasets have been utilized in Table 1 . RS and GIS techniques datasets were proven excellent for large areas and difficult terrain 30 , 31 . It was difficult to collect data manually in such an area; hence, remote detection and data collection are more appropriate and easier.

Landsat data

Landsat imagery is an efficient data source for analyzing mountain landscapes, particularly difficult terrain because of its wide and remote coverage. In this study, Landsat 4/5 Thematic Mapper (TM), 7 Enhanced Thematic Mapper Plus (ETM +) & 8 Operational Land Imager (OLI) data were utilized for the temporal analysis and vegetation indices analysis. Multiple Landsat images were downloaded from the official website of USGS ( www.usgs.gov ).

The spatial resolution of Landsat 7 is 30 × 30 m, the revisit time of the satellite is 16 days whereas, the swath width of the satellite is 185 km. Researchers used this data in different kinds of studies due to its large area coverage, fine resolution, and result accuracy, therefore in this study, Landsat data was selected to be used.

Digital elevation model

The Digital Elevation Model (DEM) is an effective mode of 3D representation of the earth’s surface. DEM with a spatial resolution of 30m is utilized for ground elevation estimation. DEM of 30-m resolution have been utilized in the study to check the elevation and Slope of the area. Elevation and slope affect the LULC of the area, therefore 30-m resolution was sufficient for the required analysis. DEM data is freely available on the official website. Data is freely available and can be downloaded from the official website of Earth Explorer.

After data collection from the various websites, data was segregated and compiled for analysis. The study area was analyzed for the twenty-five years from 1995 to 2020 and time series data was collected with a five-year gap e.g. (1995, 2000, 2005, 2010, 2015, and 2020). Scan Line error (SLE) was present in Landsat 7 (2005 & 2010 Imagery). SLE was removed from the Landsat 7 imagery via the Landsat toolbox extension.

Error handling

Due to the missing data of the Scan Line Corrector (SLC) of the Landsat instrument in 2003, approximately 20 to 22% of data went missing, and gaps were generated in the imagery. This error made the image difficult to detect. Therefore, the technique of Gap-Fill proposed by the reference 32 had been applied and error was removed from the 2007 and 2010 imagery.

Index calculation

Raw satellite data was collected freely from the website for further processing and calculation of the vegetation indices. The satellite data was converted from Digital Numbers (DN) to the values of the Top-of Atmospheric Reflectance (TOA). It was a two-step method (1) radiance of DN and (2) radiance to TOA 33 , 34 .

Due to the presence of a Scan Line Error (SLE) in Landsat 7 DN, conversion was done manually. The conversion of Landsat 7 data to DN was presented in Eq.  1 .

In this equation where Lλ was found to be the calculated radiance [in Watts per square meter  ∗  μm  ∗  ster)], DN7 was the Landsat 7 ETM + DN data, and the gain and bias were band-specific numbers 21 . Radiance to TOA conversion was also done for Landsat 7. The reflectance could be thought of as a “planetary albedo”, or a fraction of the sun’s energy that was reflected by the surface ( 22 ; Eq. ( 2 ).

Rλ = π  ∗  Lλ  ∗  d2 Esun,λ  ∗  sin (θSE) (3) where Rλ was the reflectance (unitless ratio), Lλ was the radiance calculated in Eq. ( 2 ), d was the earth-sun distance (in astronomical units), Esun, λ was the band-specific radiance emitted by the sun, π was a constant value. The process of (TOA) was done for Landsat 8 via band 10. Thermal Infra-Red DN could be converted into TOA spectral radiance utilizing rescaling factors. A Set of equations had been used for LST calculation. DN to TOA conversion was presented in equation Eq. ( 3 ).

where, Lλ = TOA spectral radiance (in Watts/(sq. meter  ∗  μm), ML = Radiance multiplicative Band No, AL = Radiance add band, Qcal = DN, Qi band correction value. The temperature was converted from Kelvin to Celsius via Eq. ( 4 ).

Finally, NDVI was calculated. The reflectance known as NDVI, a measure of greenness and proxy for vegetation degradation, was captured (Eq.  5 ).

In this equation RED = DN value of RED band and NIR = DN value of Near-Infrared band, after that land Surface Emissivity was calculated via square of NDVI. Lastly, LST was calculated for the selected study area using Eq.  6

In this equation BT = TOA, whereas, brightness in 0 C, λ = wavelength, similarly E = Land surface Emissivity, and C = velocity of light.

Demographic and socioeconomic information

The data in Table 2 shows that the surveyed population is primarily composed of males (56.7%) and females (43.3%). The percentages were calculated based on the total number of respondents.

The data in Table 3 indicates that the majority of the sample respondents fell within the 18–24 years age range (60.5%), followed by smaller proportions in older age ranges. This age distribution provides insight into the demographic composition of the surveyed population and can be useful for understanding how different age groups perceive and respond to the issues related to protecting natural resources.

The data in Table 4 provides insight into the different occupational roles present within the surveyed population. The majority of respondents were students (53.5%), followed by government/private/NGO employees (18.5%) and individuals with other occupations such as housewives, the unemployed, and those involved in agriculture, livestock farming, daily wage labor, and business. This distribution can offer context for understanding how various segments of the population view and engage with natural resource protection measures.

The data in Table 5 provides insight into the educational backgrounds of the surveyed individuals. The majority had a college/university degree (59.2%), followed by postgraduate degree holders (14.0%). Smaller proportions had secondary education, primary education, or no formal education. This distribution offers context for understanding how education levels might influence perceptions and attitudes toward natural resource protection measures.

Table 6 provides insight into the household size of the surveyed individuals. The most common household sizes were those with 5–6 people (35.0%) and 7–8 people (25.5%). It is observed that the educated respondents had more awareness about environmental degradation and the impact of anthropogenic activities. This distribution can offer context for understanding the living arrangements of the surveyed population, which might impact their resource consumption patterns and attitudes toward environmental protection.

In each group of Table 7 , the frequency and percent values indicate the number and proportion of respondents involved in the specified type of agricultural activity. This data provided insights into the agricultural practices and activities of the surveyed population, showing the prevalence of different crops and livestock within their households. This information can be useful for understanding the composition of their agricultural practices and their potential impact on the environment and natural resources.

Table 8 provides insights into the purpose of agricultural activities within the sample respondents. The majority of respondents engaged in subsistence farming (49.0%), followed by those involved in both subsistence and commercial farming (38.9%), and a smaller proportion focused on commercial farming alone (12.1%). Although subsistence farming contains a major share of the area, but still commercial farming has a second major share. Commercial farming uses fertilizers and pesticides which harm human health as well as deteriorate the environment. This distribution sheds light on the nature of agricultural practices within the surveyed community and how they might impact natural resource utilization and conservation efforts.

Table 9 revealed that more than half of the surveyed individuals were aware of the environmental issues in the Pir Chinasi National Park (52.9%), while the remaining respondents were not aware (47.1%). This awareness information was crucial for understanding the level of knowledge within the surveyed population about the environmental challenges in the region and can guide efforts to improve awareness and engagement in addressing those issues.

These percentages in Table 10 indicate the proportion of respondents who perceive each anthropogenic activity as a significant contributor to the depletion of natural resources in the area. The respondents’ opinions pointed to deforestation, road construction, livestock grazing, and tourism as the main activities causing resource depletion, with varying levels of consensus among the surveyed population. This information can be valuable for understanding the perceived drivers of environmental degradation and a reference 35 can help guide strategies for mitigating these activities to ensure sustainable resource management 36 .

Table 11 indicates the varied perceptions within the surveyed population regarding the impact of climate change on the natural resources of the Pir Chinasi National Park. Responses ranged from highly negative to highly positive, with a significant portion of respondents expressing a somewhat negative perception. These responses provide insight into the diverse perspectives on the potential consequences of climate change on the local environment and resources 31 .

The percentages in Table 12 indicate the proportion of respondents who utilize each type of fuel for cooking and heating purposes. The data provides insights into the fuel preferences within the surveyed community, which have significant implications for energy consumption patterns and their potential environmental impact. The use of firewood, LPG, natural gas, and electric heaters appears to be relatively common among the surveyed population 37 .

Table 13 reveals that a significant portion of the surveyed population did not have access to alternative sources of energy, such as solar power (65.0%), while a minority did have access (35.0%). This information provides insights into the availability and adoption of renewable energy solutions within the surveyed community 36 , 37 , which can have significant implications for energy security, environmental sustainability, and quality of life.

Table 14 provides insights into the livestock ownership patterns within the surveyed community. The majority owned 1–5 animals (43.9%), with smaller proportions owning 6–10 animals (17.2%) and 11–20 animals (7.6%). A small percentage owned more than 50 animals (1.3%), and a significant proportion did not own any livestock (29.9%). This distribution offers an understanding of the diversity in livestock ownership and its potential implications for resource use and management.

Table 15 provides insight into where the surveyed individuals’ livestock are allowed to graze. The responses indicated that animals were allowed to graze on private land (26.1%), common land (27.4%), and forests (17.2%), while a notable percentage did not have animals that graze (29.3%). Understanding the locations where livestock is grazed can help assess the potential impact on these areas and inform resource management strategies.

Table 16 reveals that a majority of the surveyed population had observed changes in the natural resources of the Pir Chinasi National Park area over the past 5–10 years (59.9%), whereas a smaller proportion (40.1%) has not perceived such changes. This information provides insights into the perceived dynamics of natural resource changes in the region and can help in understanding the evolving environmental conditions and potential factors contributing to these changes.

The percentages in Table 17 indicate the proportion of respondents who believed that each measure should be taken to protect the natural resources in the area. It appears that reforestation programs had the highest agreement (58.0%), followed by strict enforcement of environmental protection laws (53.2%). Other measures such as public awareness campaigns, collaboration with local communities, sustainable tourism practices, and sustainable agriculture practices also received notable support from the surveyed population. These opinions provide insight into the potential strategies that could be pursued to safeguard the natural resources in the Pir Chinasi National Park area according to the surveyed community.

Time series analysis of vegetation change dynamics of the selected area was evaluated via NDVI. Spatial and temporal dynamics of vegetation degradation, depending on the time scale were analyzed. The status of the landscape was reflected in vegetation indices (VIs), and their interpretation across time explained patterns in vegetative greening (land improvement) and browning (land deterioration). Based on a long-term investigation of vegetation dynamics, Figs. 2 – 4 illustrate swings between periods of gradual deterioration and minor improvements (greening). A slightly declining trend of NDVI was observed during the study period (1995–2020). During the entire study period mean NDVI (maximum = 0.28 (1995)/minimum − 2.8 (2010) was identified see Fig.  2 for instance. In 2000, mean NDVI was monitored at 0.24 (Fig.  2 b). Similarly, the NDVI of 2005 was also calculated via the set of formulas mentioned in the materials and method Section 38 , 39 , 40 .

shows the satellite-based NDVI variation temporally from 1995 to 2020.

Slope patterns of the study area.

Land Surface Temperature (LST) ° C trends of the study area during the study period. The map was created using ArcGIS Desktop software (version 10.8.2, Esri lnc., Redlands,CA, USA). ArcGIS is widely used for creating slope maps from Digital Elevation Models (DEMs).

The mean value of NDVI was 0.22 to 0.76. Pir Chinasi National Park was known for tourism, this decline can be attributed to the expansion of built-up areas to accommodate growing tourist demands, leading to habitat loss and ecological changes. This decline can be attributed to the expansion of built-up areas to accommodate growing tourist demands, leading to habitat loss and ecological changes. NDVI for the year 2015 and 2020 was observed 0.07–0.53 to 0.08–0.57 (Fig.  2 e and f).

Land surface temperature (LST)

Surface temperature had a significant impact on the local ecology and biodiversity. Therefore, in this study, LST for the study period was also calculated to investigate the impacts of temperature on vegetation. The temperature was estimated via Landsat imagery for the selected images from 1995 to 2020 (Fig.  4 ). The temperature range for 1995 was 19 minimum to 29 o C maximum (Fig.  4 a), 23° to 36° for 2000 (Fig.  4 b), whereas 21° to 35° temperature range was observed for 2005 (Fig.  4 c). 29 °C to 39 °C for 2010 see Fig.  4 d. For the year 2015 and 2020 (9°–26 °C and 17–37 °C) were observed respectively.

It has tools specifically for slope calculation, where the slope can be expressed in degrees or as a percentage. Slope analysis, spatial analysis, and 3D visualization using extensions like Spatial Analyst or 3D Analyst. User-friendly interface, extensive documentation, and strong support for various geospatial analyses.

Ground truthing

Data collection from the field was time-consuming due to the monetary constraints 23 , 24 , 33 , therefore, Google Earth data was utilized for the investigation of urbanization patterns. Deforestation was observed where recreational activities were occurring, see Fig.  5 .

Google Earth image of Pir Chinasi National Park.

Figure  5 highlighted the ground truthing based on detailed information about vegetation to validate the remote sensing. The figure shows urbanization patterns. Ground truthing with physical sampling strengthened the validation of the NDVI analysis. Map was generated using the open-source software (QGIS 3.8) along with that an open-source software Google Earth Pro (GEP) was utilized for image data collection. KML files were generated using GEP and converted into shapefiles in the QGIS environment. The geographical coordinates (34°23′22"N 73°32′57"E) were used. https://earth.google.com/web/search/Pir+Chinasi/@34.38987286,73.55007956,2825.08631358a,2608.57706282d,35y,0h,0t,0r/data=CigiJgokCXkkdsjBB0FAEWoFBBCW-EBAGZ7DNELhfFJAIah7yxvnaFJAOgMKATA

Numerous reports have been published relating NDVI to rainfall patterns. In a similar vein, by reference 41 discovered that precipitation in Ethiopia’s semi-arid areas was strongly correlated with NDVI levels throughout the growing season. According to the reference 42 , 80% of the Caatinga vegetation productivity anomaly may be accounted for by the rainfall anomaly using the vegetation index as a proxy taken by reference 43 in the Horn of Africa reported similar findings.

In China, a significant link between NDVI and precipitation was discovered by researchers 17 , 44 . According to reference 45 , NDVI across Northern Mongolia showed a positive association with yearly precipitation and a negative correlation with temperature 16 . Researchers 13 , 46 , 47 , found that NDVI was more closely connected with rainfall than temperature on the Tibetan Plateau and the Loess Plateau, respectively.

In actuality, when evaluating vegetation response to climate change the perceptions of the residents can also be taken 48 . In this study, the respondents perceived a major change in vegetation due to climate change and anthropogenic activities. A weak or negative (−) NDVI-Rainfall link implies that vegetation dynamics are influenced by the climate; a robust and positive ( +) connection indicates changes in vegetation caused by humans. The researcher 49 also found comparable results in its worldwide survey.

According to researcher 50 , precipitation is the key environmental element influencing NDVI changes in the Heihe River Basin, China. According to another researcher 51 , precipitation and land cover in Kenya’s Mara River Basin revealed a highly positive association. According to reports, the primary human activities responsible for the deterioration of the local vegetation and the disappearance of the forest in this region include grazing, agriculture-related settlement growth, and wood collecting 52 , 53 . According to researcher 51 , the highlands’ growing pace of forest degradation is a sign of how much the high population density, 350inch/km 2 2 , 24 , 39 value and exploit this forest on an economic, social, ethnological, and cultural level. It has been studied how some of the aforementioned factors relate to NDVI. The NDVI of the Wei and Jing River Basins, China, was shown to be primarily influenced by three variables: temperature, soil moisture, and precipitation. The researcher 54 demonstrated a tight link between the dynamics of grassland and temperature using NDVI-max at an annual time scale. Future studies could focus on exploring the connection between the environmental elements and the dynamics of the vegetation on the Pir Chinasi National Park.

Conclusions

The study described the dynamics of vegetation and how it responded to changing rainfall patterns and human pressures. It was proven that vegetation change dynamics and degradation are impacted by climate change as indicated by long-term rainfall anomaly (RAI). The study found that 10.81% of the area experienced substantial deterioration, while 23.57% experienced minor degradation. The total area of degraded lands was 34.38% suggesting a major change. To overcome these phenomena, steps of adaptation by the public–private sector are necessary. Improvement in plant cover was observed which is a positive sign. According to the NDVI-Rainfall relationships, the area was found to be significantly impacted by human pressures which require control by public–private partnership. The area under the influence of climatic variability and change is also large enough which requires immediate attention. Sustainable land management practices of conservation tillage, integrated pest management, and agroforestry help preserve soil health, water quality, and biodiversity while reducing erosion, pollution, and the degradation of natural resources. landscape restoration projects of reforestation, wetland restoration, soil erosion control, and the removal of invasive species are essential to achieve land degradation neutrality at the watershed scale. The quality and health of the vegetation were further harmed by unsustainable human activities such as agricultural development, overgrazing, settlement growth, and wood exploitation. However, climate change proved to be a driver of these phenomena. Even though these were the major causes, the degree to which the flora on the plateau tends to green up or brown out, could also be influenced by environmental variables such as soil moisture, solar radiation intensity and duration, drainage density, and topographic restrictions. This study does not conclude that the sole variables affecting vegetation dynamics are rainfall and human influences.

In areas of the park where vegetation patterns indicate deterioration or loss of forest cover, afforestation, and replanting efforts could be done. These initiatives have the potential to improve ecosystem resilience to climate change, restore biodiversity, and sequester carbon. Implement sustainable land management techniques that boost agricultural production, and preserve and improve plant cover, such as conservation agriculture, rotational grazing, and agroforestry. These methods could lessen soil erosion, increase water retention, and protect the benefits provided by ecosystem services. In response to climate change and human intervention establish buffer zones, wildlife corridors, and protected areas to preserve important ecosystems and enhance landscape connectivity.

Data availability

Data is provided within the manuscript or supplementary information files.

Guan, X., Liu, M. & Meng, Y. A comprehensive ecological compensation indicator based on pollution damage—Protection bidirectional model for river basin. Ecol. Indic. 126 , 107708 (2021).

Article   Google Scholar  

Fu, B. et al. Current condition and protection strategies of Qinghai-Tibet Plateau ecological security barrier. Bull. Chinese Acad. Sci. 36 , 1298–1306 (2021).

Google Scholar  

Wen, Z., Wu, S., Chen, J. & Lü, M. NDVI indicated long-term interannual changes in vegetation activities and their responses to climatic and anthropogenic factors in the three gorges reservoir region, China. Sci. Total Environ. 574 , 947–959 (2017).

Article   ADS   CAS   PubMed   Google Scholar  

Shiweda, M., Shivute, F., Sales, A. R. & Pereira, M. J. Climate change and anthropogenic factors are influencing the loss of habitats and emerging human-elephant conflict in the Namib desert. Sustainability 15 , 12400 (2023).

Du, Y. et al. Human activity coupled with climate change strengthens the role of lakes as an active pipe of dissolved organic matter. Earth’s Futur. https://doi.org/10.1029/2022EF003412 (2023).

Gabriele, M., Brumana, R., Previtali, M. & Cazzani, A. A combined GIS and remote sensing approach for monitoring climate change-related land degradation to support landscape preservation and planning tools: The Basilicata case study. Appl. Geomat. 15 , 497–532 (2023).

Subramanian, A. et al. Long-term impacts of climate change on coastal and transitional eco-systems in India: An overview of its current status, future projections, solutions, and policies. RSC Adv. 13 , 12204–12228 (2023).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Zheng, K. et al. Impacts of climate change and anthropogenic activities on vegetation change: Evidence from typical areas in China. Ecol. Indic. 126 , 107648 (2021).

Du, J. et al. Spatiotemporal changes in vegetation coverage and its causes in China since the Chinese economic reform. Environ. Sci. Pollut. Res. 27 , 1144–1159 (2020).

Lamchin, M. et al. Long-term trend of and correlation between vegetation greenness and climate variables in Asia based on satellite data. MethodsX 5 , 803–807 (2018).

Article   PubMed   PubMed Central   Google Scholar  

Pettorelli, N. et al. Using the satellite-derived NDVI to assess ecological responses to environmental change. Trends Ecol. Evol. 20 , 503–510 (2005).

Article   PubMed   Google Scholar  

Sharon, E. & Nicholson, M. L. D. A comparison of the vegetation response to rainfall in the sahel and east africa, using normalized difference vegetation from NOAA AVHRR. Clim. Change 21 , 99–104 (1990).

Chu, D., Lu, L. & Zhang, T. Sensitivity of normalized difference vegetation index (NDVI) to seasonal and interannual climate conditions in the Lhasa area, Tibetan Plateau, China. Arctic Antarct. Alp. Res. 39 , 635–641 (2007).

Article   ADS   Google Scholar  

Zawude Bakure, B., Hundera, K. & Abara, M. Review on the effect of climate change on ecosystem services. IOP Conf. Ser. Earth Environ. Sci. 1016 , 012055 (2022).

Salimon, C. & Anderson, L. How strong is the relationship between rainfall variability and caatinga productivity? A case study under a changing climate. An. Acad. Bras. Cienc. 90 , 2121–2127 (2018).

Nanzad, L. et al. NDVI anomaly for drought monitoring and its correlation with climate factors over Mongolia from 2000 to 2016. J. Arid Environ. 164 , 69–77 (2019).

Chen, H. et al. The impacts of climate change and human activities on biogeochemical cycles on the Qinghai-Tibetan Plateau. Glob. Chang. Biol. 19 , 2940–2955 (2013).

Article   ADS   PubMed   Google Scholar  

Liu, Z., Liu, Y. & Li, Y. Anthropogenic contributions dominate trends of vegetation cover change over the farming-pastoral ecotone of northern China. Ecol. Indic. 95 , 370–378 (2018).

Zhang, B. et al. Spatial distribution and suitability evaluation of nighttime tourism in Kunming utilizing multi-source data. Heliyon 9 , e16826 (2023).

Liu, Y., Zhang, M. & Liu, R. The impact of income inequality on carbon emissions in china: A household-level analysis. Sustain 12 , 1–22 (2020).

XU, J. et al. The melting Himalayas: Cascading effects of climate change on water, biodiversity, and livelihoods. Conserv. Biol. 23 , 520–530 (2009).

Article   CAS   PubMed   Google Scholar  

Hussein, A. J. & Salih, S. A. Quantifying the reliability of volumetric and areal calculation with UAV-generated DEMs: A comparative study with ground truth data. Tikrit J. Eng. Sci. 31 , 43–55 (2024).

Sirianni, H., Richter, J., Sirianni, M. J. & Pettyjohn, S. Shoreline classification maps and ground truth data for the Neuse River Estuary North Carolina. Sci. Data 11 , 103 (2024).

Kraatz, S. et al. Comparing NISAR (Using Sentinel-1), USDA/NASS CDL, and ground truth crop/Non-crop areas in an urban agricultural region. Sensors 23 , 8595 (2023).

Article   ADS   PubMed   PubMed Central   Google Scholar  

Zoungrana, B. J. B. Vegetation dynamics in the southwest of Burkina Faso in response to rainfall variability and land use. Thesis 162 (2016).

Musau, J., Patil, S., Sheffield, J. & Marshall, M. Vegetation dynamics and responses to climate anomalies in East Africa. Earth Syst. Dyn. Discuss. 2018 , 1–27 (2018).

Chinasi, P. Blog. 2024 https://eagleeye.com.pk/pttl/pir-chinasi/ .

Ministry of Finance. Economic Survey of Pakistan 2018–19. Chapter 14- Energy 183–196 (2019).

Sharma, N., Behera, M. D., Das, A. P. & Panda, R. M. Plant richness pattern in an elevation gradient in the Eastern Himalaya. Biodivers. Conserv. 28 , 2085–2104 (2019).

Ahmed, N. Spatial vulnerability assessment of urbanites of Rawalpindi to the combined effects of climate change and pandemics using GIS mapping. Local Environ. https://doi.org/10.1080/13549839.2024.2353048 (2024).

Ahmed, N., Padda, I. U. H. & Khan, A. Mapping adaptive capacity of urban residents: a case study of Rawalpindi Pakistan. NUST J. Soc. Sci. Humanit. 8 , 151–169 (2023).

Muir, C., Southworth, J., Khatami, R., Herrero, H. & Akyapı, B. Vegetation dynamics and climatological drivers in ethiopia at the turn of the century. Remote Sens. 13 , 3267 (2021).

Puente, M. & Vélez, V. N. Ground-truthing as critical race Feminista methodology: Toward an embodied and community-centered GIS in educational inquiry. Int. J. Qual. Stud. Educ. 37 , 1287–1306 (2024).

Baltagi, B. H. Random effects and spatial autocorrelation with equal weights. Econ. Theory 22 , 973–984 (2006).

Article   MathSciNet   Google Scholar  

Ahmed, N. et al. Climate change adaption strategies in urban communities: New evidence from Islamabad Pakistan. Environ. Sci. Pollut. Res. https://doi.org/10.1007/s11356-023-25316-x (2023).

Ahmed, N. & Luqman, M. Explaining urban communities’ adaptation strategies for climate change risk: Novel evidence from Rawalpindi Pakistan. Nat. Hazards https://doi.org/10.1007/s11069-024-06501-8 (2024).

Batool, Z., Ahmed, N. & Luqman, M. Examining the role of ICT, transportation energy consumption, and urbanization in CO2 emissions in Asia: A threshold analysis. Environ. Sci. Pollut. Res. 30 , 78482–78494 (2023).

Article   CAS   Google Scholar  

Bedair, H. et al. Global warming status in the african continent: Sources, challenges, policies, and future direction. Int. J. Environ. Res. 17 , 45 (2023).

Gajendiran, K., Kandasamy, S. & Narayanan, M. Influences of wildfire on the forest ecosystem and climate change: A comprehensive study. Environ. Res. 240 , 117537 (2024).

Garoon, E. M. & Mule, M. B. A multiscale analysis and classification of normalized difference vegetation index value method in Kannad Taluka, Aurangabad District, using remote sensing and GIS. J. Environ. Biol. 44 , 335–344 (2023).

Zewdie, W. & Csaplovics, E. Evaluation of rainfall and NDVI anomalies using distributed lag models. in Algorithms and Technologies for Multispectral, Hyperspectral, and Ultraspectral Imagery XXI (eds. Velez-Reyes, M. & Kruse, F. A.) vol. 9472 94721O (2015).

Andersen, A. N. & Vasconcelos, H. L. Historical biogeography shapes functional ecology: Inter-continental contrasts in responses of savanna ant communities to stress and disturbance. J. Biogeogr. 49 , 590–599 (2022).

Ghebrezgabher, M. G., Yang, T., Yang, X. & Eyassu Sereke, T. Assessment of NDVI variations in responses to climate change in the Horn of Africa Egypt. J. Remote Sens. Sp. Sci. 23 , 249–261 (2020).

Chen, D. How visitors perceive heritage value—A quantitative study on visitors’ perceived value and satisfaction of architectural heritage through SEM. Sustainability 15 , 9002 (2023).

Enebish, B., Dashkhuu, D., Renchin, M., Russell, M. & Singh, P. Impact of climate on the NDVI of northern Mongolia. J. Indian Soc. Remote Sens. 48 , 333–340 (2020).

Hao, Y. et al. States, trends, and future of aquaponics research. Sustainability 12 , 7783 (2020).

Ning, T., Liu, W., Lin, W. & Song, X. NDVI variation and its responses to climate change on the Northern loess plateau of China from 1998 to 2012. Adv. Meteorol. 2015 , 1–10 (2015).

Sapkale, J. B., Mane, M. M., Susware, N. K. & Sapkale, S. J. Dynamic changes in coastal geomorphology of Shiroda coasts, using remote sensing and GIS: An approach to climate change and coastal disaster risk. Disaster Adv. 16 , 20–32 (2023).

Bai, Z. G., Dent, D. L., Olsson, L. & Schaepman, M. E. Proxy global assessment of land degradation. Soil Use Manag. 24 , 223–234 (2008).

Yuan, L., Chen, X., Wang, X., Xiong, Z. & Song, C. Spatial associations between NDVI and environmental factors in the Heihe River Basin. J. Geogr. Sci. 29 , 1548–1564 (2019).

Mngube, F. M., Kapiyo, R., Aboum, P., Anyona, D. & Dida, G. O. Subtle impacts of temperature and rainfall patterns on land cover change overtime and future projections in the Mara river Basin, Kenya. Open J. Soil Sci. 10 , 327–358 (2020).

Rawat, A., Kumar, D. & Khati, B. S. A review on climate change impacts, models, and its consequences on different sectors: A systematic approach. J. Water Clim. Chang. 15 , 104–126 (2024).

Vinet, L. & Zhedanov, A. A ‘missing’ family of classical orthogonal polynomials. J. Phys. A Math. Theor. 44 , 1–22 (2011).

Huang, S. et al. A novel fusion method for generating surface soil moisture data with high accuracy, high spatial resolution, and high Spatio-temporal continuity. Water Resour. Res. 58 , 1–26 (2022).

Download references

Acknowledgements

The authors extend their appreciation to Researchers Supporting Project number (RSP-2024 R369), King Saud University, Riyadh, Saudi Arabia.

Author information

Authors and affiliations.

Department of Environmental Science, International Islamic University, Islamabad, Pakistan

Muhammad Akhlaq Farooq & Muhammad Asad Ghufran

Department of Economics, National University of Modern Languages, Islamabad, Pakistan

Naeem Ahmed & Aamir Amanat

Department of Biochemistry, College of Science, King Saud University, P.O. Box 2455, 11451, Riyadh, Saudi Arabia

Kotb A. Attia & Arif Ahmed Mohammed

Kafr-El-Sheikh University, Kafr el-Sheikh, 33516, Egypt

Yaser M. Hafeez

Rice Research Program, Crop Sciences Institute (CSI), National Agricultural Research Centre (NARC), Park Road 44000, Islamabad, Pakistan

Muhammad Shahbaz Farooq

National Institute for Genomics and Advanced Biotechnology (NIGAB), National Agricultural Research Centre (NARC), Park Road 45500, Islamabad, Pakistan

Muhammad Uzair

COMSATS University, Islamabad, Pakistan

You can also search for this author in PubMed   Google Scholar

Contributions

Conceptualization, Muhammad Akhlaq Farooq and Naeem Ahmed; Data curation, Muhammad Akhlaq Farooq and Muhammad Shahbaz Farooq; Formal analysis, Muhammad Akhlaq Farooq, Muhammad Ghufran, Aamir Amanat, and Saima Naz; Funding acquisition, Naeem Ahmed, Muhammad Shahbaz Farooq, and Muhammad Uzair; Investigation, Muhammad Ghufran, Yasser M Hafez, and Aamir Amanat; Methodology, Aamir Amanat and Saima Naz; Project administration, Naeem Ahmed; Resources, Naeem Ahmed, Muhammad Uzair and Saima Naz; Software, Aamir Amanat; Supervision, Naeem Ahmed; Validation, Aamir Amanat; Visualization, Muhammad Akhlaq Farooq, Kotb A. Attia, Arif Ahmed Mohammed; Writing – original draft, Muhammad Akhlaq Farooq; Writing – review & editing, Muhammad Akhlaq Farooq, Muhammad Ghufran, Naeem Ahmed, Muhammad Shahbaz Farooq, Muhammad Uzair, Kotb A. Attia, Arif Ahmed Mohammed and Saima Naz. All authors have read and agreed to the published version of the manuscript.

Corresponding authors

Correspondence to Naeem Ahmed or Muhammad Uzair .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Research Ethics

Experimental research and field studies on plants (either cultivated or wild), including the collection of plant material, complied with relevant institutional, national, and international guidelines and legislation. Prior approval was undertaken from the Offices of Research, Innovation and Commercialization, National University of Modern Languages, Islamabad-Pakistan. We also took appropriate permission from the farm or field owner during specimens’ collection and experimentation. We confirm that during the collection and execution of the experiment, the authors have complied with the IUCN Statement on Research Involving Species at Risk of Extinction and the Convention on the Trade in Endangered Species of Wild Fauna and Flora. All methods were performed in accordance with the relevant guidelines and regulations. All experimental protocols were approved by Offices of Research, Innovation and Commercialization, National University of Modern Languages, Islamabad-Pakistan.

Additional information

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 licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, 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 you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. 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-nc-nd/4.0/ .

Reprints and permissions

About this article

Cite this article.

Farooq, M.A., Ghufran, M.A., Ahmed, N. et al. Remote sensing analysis of spatiotemporal impacts of anthropogenic influence on mountain landscape ecology in Pir Chinasi national park. Sci Rep 14 , 20695 (2024). https://doi.org/10.1038/s41598-024-71689-5

Download citation

Received : 05 December 2023

Accepted : 30 August 2024

Published : 05 September 2024

DOI : https://doi.org/10.1038/s41598-024-71689-5

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

• Landsat NDVI
• Spatiotemporal analysis
• Land surface temperature
• Anthropogenic
• Overgrazing
• Pir-Chinasi national park

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

Sign up for the Nature Briefing: Anthropocene newsletter — what matters in anthropocene research, free to your inbox weekly.