Application of multivariate statistical techniques in the assessment of groundwater quality in seawater intrusion area in Bafra Plain, Turkey

Multivariate statistical techniques such as cluster analysis and principal component analysis were performed on 28 groundwater wells in Bafra Plain. Cluster analysis results show that the groundwater in the study area is classified into three groups (A, B, and C), and factor analysis indicates that groundwater is composed of 89.64 % of total variance of 12 variables and is mainly affected by three factors. Factor 1 (seawater salinization) includes concentrations of electrical conductivity, TDS, Cl^?, Na⁺, and sodium adsorption ratio, factor 2 (mixing water) includes δ¹⁸O, δD, and T, and factor 3 (fresh) includes Ca²⁺. For determination of the source of water, Ca/Cl, Cl/HCO₃, Mg/Cl, and Ca/Na as initials and Mg/Ca and SO₄/Cl as molar rates which were identified, the rates had been found to be very useful. Cluster analysis was made by using these rates and the waters were classified in two groups (group 1 and group 2). First group waters were affected by seawater, and the second group were very less affected by freshwater or seawater. According to the comparison of two different parameters, group 1 comprised group A and group B-2, -3, and -4 from the same wells, and group 2 comprised group B-1 and group C from the same well. As a result of this study, it could be said that multivariate statistical methods gave very useful results for the determination of the source.     

Using multivariate statistical methods to assess the groundwater quality in an arsenic-contaminated area of Southwestern Taiwan

Groundwater is a major water resource in Southwestern Taiwan; hence, long-term monitoring of water quality is essential. The study aims to assess the hydrochemical characteristics of water in the arsenic-contaminated aquifers of Choushui River alluvial fan and Chianan Plain, Taiwan using multivariate statistical methods, namely, factor analysis (FA), cluster analysis (CA), and discriminant analysis (DA). Factor analysis is applied to reveal the processes controlling the hydrochemistry of groundwater. Cluster analysis is applied to spatially categorize the collected water samples based on the water quality. Discriminant analysis is then applied to elucidate key parameters associated with the occurrence of elevated As concentration (>10 μg L(-1)) in groundwater. Major water types are characterized as Na-Ca-Cl and Na-Mg-Cl in the Choushui River alluvial fan and Chianan Plain, respectively. Inorganic species of arsenic (As), particularly As(III), prevail in these two groundwater catchments, and their levels are higher in the Chianan Plain than in the Choushui River alluvial fan. Through FA, three factors, namely, the degree of salination, As reduction, and iron (Fe) reduction, are determined and denoted irrespective of some differences between the factorial compositions. Spatial distribution patterns of factors As reduction and Fe reduction imply that the redox zonation is delineated by As- and Fe-dominance zones separately. The results of CA demonstrate that three main groups can be properly explained by the factors extracted via FA. Three- (Fe(2+), Fe(3+), and NH (4) (+) ) and four-parameters (Fe(2+), Fe(3+), NH (4) (+) , and Ca(2+)) derived from discriminant analysis for Choushui River alluvial fan and Chianan Plain are elucidated as key parameters affecting the distribution of As-contained groundwater. The analytical results indicate that the reductive dissolution of Fe minerals is prerequisite for the mobilization of As, whereas the shift of redox condition from Fe- to As-reducing leads to the accumulation of dissolved As in this area.     

Water quality assessment and source identification of Daliao river basin using multivariate statistical methods

Multivariate statistical methods, such as cluster analysis (CA), discriminant analysis (DA) and principal component analysis (PCA), were used to analyze the water quality dataset including 13 parameters at 18 sites of the Daliao River Basin from 2003-2005 (8424 observations) to obtain temporal and spatial variations and to identify potential pollution sources. Using Hierarchical CA it is classified 12 months into three periods (first, second and third period) and the 18 sampling sites into three groups (groups A, B and C). Six significant parameters (temperature, pH, DO, BOD(5), volatile phenol and E. coli) were identified by DA for distinguishing temporal or spatial groups, with close to 84.5% correct assignment for temporal variation analysis, while five parameters (DO, NH(4)(+)-N, Hg, volatile phenol and E. coli) were discovered to correctly assign about 73.61% for the spatial variation analysis. PCA is useful in identifying five latent pollution sources for group B and C (oxygen consuming organic pollution, toxic organic pollution, heavy metal pollution, fecal pollution and oil pollution). During the first period, sites received more oxygen consuming organic pollution, toxic organic pollution and heavy metal pollution than those in the other two periods. For group B, sites were mainly affected by oxygen consuming organic pollution and toxic organic pollution during the first period. The level of pollution in the second period was between the other two periods. For group C, sites were mainly affected by oil pollution during the first period and oxygen consuming organic pollution during the third period. Furthermore, source identification of each period for group B and group C provided useful information about seasonal pollution. Sites were mainly affected by fecal pollution in the third period for group B, indicating the character of non-point source pollution. In addition, all the sites were also affected by physical-chemistry pollution. In the second and third period for group B and second period for group C sites were also affected by natural pollution.     

Application of multivariate statistical techniques in the assessment of surface water quality in Uluabat Lake, Turkey

The application of different multivariate statistical approaches for the interpretation of a complex data matrix obtained during the period 2004-2005 from Uluabat Lake surface water is presented in this study. The dataset consists of the analytical results of a 1 year-survey conducted in 12 sampling stations in the Lake. Twelve parameters (T, pH, DO, [Formula: see text], NH(4)-N, NO(2)-N, NO(3)-N, [Formula: see text], BOD, COD, TC, FC) were monitored in the sampling sites on a monthly basis (except December 2004, January and February 2005, a total of 1,296 observations). The dataset was treated using cluster analysis, principle component analysis and factor analysis on principle components. Cluster analysis revealed two different groups of similarities between the sampling sites, reflecting different physicochemical properties and pollution levels in the studied water system. Three latent factors were identified as responsible for the data structure, explaining 77.35% of total variance in the dataset. The first factor called the microbiological factor explained 32.34% of the total variance. The second factor named the organic-nutrient factors explained 25.46% and the third factor called physicochemical factors explained 19.54% of the variances, respectively.     

Application of multivariate statistical techniques in the assessment of water quality in the Southwest New Territories and Kowloon, Hong Kong

The application of different multivariate statistical techniques for the interpretation of a complex data matrix obtained during 2000?C2007 from the watercourses in the Southwest New Territories and Kowloon, Hong Kong was presented in this study. The data set consisted of the analytical results of 23 parameters measured monthly at 16 different sampling sites. Hierarchical cluster analysis grouped the 12 months into two periods and the 16 sampling sites into three groups based on similarity in water quality characteristics. Discriminant analysis (DA) provided better results both temporally and spatially. DA also offered an important data reduction as it only used four parameters for temporal analysis, affording 84.2% correct assignations, and eight parameters for spatial analysis, affording 96.1% correct assignations. Principal component analysis/factor analysis identified four latent factors standing for organic pollution, industrial pollution, nonpoint pollution, and fecal pollution, respectively. KN1, KN4, KN5, and KN7 were greatly affected by organic pollution, industrial pollution, and nonpoint pollution. The main pollution sources of TN1 and TN2 were organic pollution and nonpoint pollution, respectively. Industrial pollution had high effect on TN3, TN4, TN5, and TN6.     

Use of water quality index and multivariate statistical techniques for the assessment of spatial variations in water quality of a small river

Rapid urban development has led to a critical negative impact on water bodies flowing in and around urban areas. In the present study, 25 physiochemical and biological parameters have been studied on water samples collected from the entire section of a small river originating and ending within an urban area. This study envisaged to assess the water quality status of river body and explore probable sources of pollution in the river. Weighted arithmetic water quality index (WQI) was employed to evaluate the water quality status of the river. Multivariate statistical techniques namely cluster analysis (CA) and principal component analysis (PCA) were applied to differentiate the sources of variation in water quality and to determine the cause of pollution in the river. WQI values indicated high pollution levels in the studied water body, rendering it unsuitable for any practical purpose. Cluster analysis results showed that the river samples can be divided into four groups. Use of PCA identified four important factors describing the types of pollution in the river, namely (1) mineral and nutrient pollution, (2) heavy metal pollution, (3) organic pollution, and (4) fecal contamination. The deteriorating water quality of the river was demonstrated to originate from wide sources of anthropogenic activities, especially municipal sewage discharge from unplanned housing areas, wastewater discharge from small industrial units, livestock activities, and indiscriminate dumping of solid wastes in the river. Thus, the present study effectively demonstrates the use of WQI and multivariate statistical techniques for gaining simpler and meaningful information about the water quality of a lotic water body as well as to identify of the pollution sources.     

Identification of mangrove water quality by multivariate statistical analysis methods in Pondicherry coast,India

Different multivariate statistical analysis such as, cluster analysis, principal component analysis, and multidimensional scale plot were employed to evaluate the trophic status of water quality for four monitoring stations. The present study was carried out to determine the physicochemical parameters of water and sediment characteristics of Pondicherry mangroves—southeast coast of India, during September 2008–December 2010. Seasonal variations of different parameters investigated were as follows: salinity (10.26–35.20 psu), dissolved oxygen (3.71–5.33 mg/L), pH (7.05–8.36), electrical conductivity (26.41–41.33 ms⁻¹), sulfide (1.98–40.43 mg/L), sediment texture sand (39.54–87.31%), silt (9.89–32.97%), clay (3.06–31.20%), and organic matter (0.94–4.64%). pH, temperature, salinity, sand, silt, clay, and organic matter indicated a correlation at P < 0.01. CA grouped the four seasons in to four groups (pre-monsoon, monsoon, post-monsoon, summer) and the sampling sites in to three groups. PCA identified the spatial and temporal characteristics of trophic stations and showed that the water quality was worse in stations 3 and 4 in the Pondicherry mangroves.     

The use of multivariate statistical methods for optimization of the surface water quality network monitoring in the Paraopeba river basin,Brazil

This study sought to evaluate and propose adjustments to the water quality monitoring network of surface freshwaters in the Paraopeba river basin (Minas Gerais, Brazil), using multivariate statistical methods. A total of 13,560 valid data were analyzed for 19 water quality parameters at 30 monitoring sites, over a period of 5 years (2008–2013). The cluster analysis grouped the monitoring sites in eight groups based on similarities of water quality characteristics. This analysis made it possible to detect the most relevant monitoring stations in the river basin. The principal components analysis associated with non-parametric tests and the analysis of violation of the standards prescribed by law, allowed for identifying the most relevant parameters which must be maintained in the network (thermotolerant coliforms, total manganese, and total phosphorus). The discharge of domestic sewage and industrial wastewater, that from mining activities and diffuse pollution from agriculture and pasture areas are the main sources of pollution responsible for the surface water quality deterioration in this basin. The BP073 monitoring site presents the most degraded water quality in the Paropeba river basin. The monitoring sites BP094 and BP092 are located geographically close and they measure similar water quality, so a possible assessment of the need to maintain only one of the two in the monitoring network is suggested. Therefore, multivariate analyses were efficient to assess the adequacy of the water quality monitoring network of the Paraopeba river basin, and it can be used in other watersheds.     

Regional assessment of groundwater quality for drinking purpose

Owing to limited surface water during a long-term drought, this work attempted to locate clean and safe groundwater in the Choushui River alluvial fan of Taiwan based on drinking-water quality standards. Because aquifers contained several pollutants, multivariate indicator kriging (MVIK) was adopted to integrate the multiple pollutants in groundwater based on drinking- and raw-water quality standards and to explore spatial uncertainty. According to probabilities estimated by MVIK, safe zones were determined under four treatment conditions—no treatment; ammonium–N and iron removal; manganese and arsenic removal; and ammonium–N, iron, manganese, and arsenic removal. The analyzed results reveal that groundwater in the study area is not appropriate for drinking use without any treatments because of high ammonium–N, iron, manganese, and/or arsenic concentrations. After ammonium–N, iron, manganese, and arsenic removed, about 81.9–94.9% of total areas can extract safe groundwater for drinking. The proximal-fan, central mid-fan, southern mid-fan, and northern regions are the excellent locations to pump safe groundwater for drinking after treatment. Deep aquifers of exceeding 200 m depth have wider regions to obtain excellent groundwater than shallow aquifers do.     

Application of water quality index for groundwater quality assessment: Thirumanimuttar sub-basin, Tamilnadu, India

An attempt has been made to understand the hydrogeochemical parameters to develop water quality index in Thirumanimuttar sub-basin. A total of 148 groundwater samples were collected and analyzed for major cations and anions. The domination of cations and anions was in the order of Na>Mg>Ca>K for cations and Cl>HCO₃ >SO₄ in anions. The hydrogeochemical facies indicate alkalis (Na and K) exceed alkaline earths (Ca and Mg) and strong acids (Cl and SO₄) exceed weak acid (HCO₃). Water quality index rating was calculated to quantify overall water quality for human consumption. The PRM samples exhibit poor quality in greater percentage when compared with POM due to effective leaching of ions, over exploitation of groundwater, direct discharge of effluents and agricultural impact. The overlay of WQI with chloride and EC correspond to the same locations indicating the poor quality of groundwater in the study area. SAR, Na%, and TH were noted higher during both the seasons indicating most of the groundwater locations not suitable for irrigation purposes.     

Use of multivariate indicator kriging methods for assessing groundwater contamination extents for irrigation

Multivariate geostatistical approaches have been applied extensively in characterizing risks and uncertainty of pollutant concentrations exceeding anthropogenic regulatory limits. Spatially delineating an extent of contamination potential is considerably critical for regional groundwater resources protection and utilization. This study used multivariate indicator kriging (MVIK) to determine spatial patterns of contamination extents in groundwater for irrigation and made a predicted comparison between two types of MVIK, including MVIK of multiplying indicator variables (MVIK-M) and of averaging indicator variables (MVIK-A). A cross-validation procedure was adopted to examine the performance of predicted errors, and various probability thresholds used to calculate ratios of declared pollution area to total area were explored for the two MVIK methods. The assessed results reveal that the northern and central aquifers have excellent groundwater quality for irrigation use. Results obtained through a cross-validation procedure indicate that MVIK-M is more robust than MVIK-A. Furthermore, a low ratio of declared pollution area to total area in MVIK-A may result in an unrealistic and unreliable probability used to determine extents of pollutants. Therefore, this study suggests using MVIK-M to probabilistically determine extents of pollutants in groundwater.     

Evaluation of spatial and temporal variations in marine sediments quality using multivariate statistical techniques

Spatial and temporal variations of sediment quality in Matanzas Bay (Cuba) were studied by determining a total of 12 variables (Zn, Cu, Pb, As, Ni, Co, Al, Fe, Mn, V, CO₃ ^2?, and total hydrocarbons (THC). Surface sediments were collected, annually, at eight stations during 2005–2008. Multivariate statistical techniques, such as principal component (PCA), cluster (CA), and lineal discriminant (LDA) analyses were applied for identification of the most significant variables influencing the environmental quality of sediments. Heavy metals (Zn, Cu, Pb, V, and As) and THC were the most significant species contributing to sediment quality variations during the sampling period. Concentrations of V and As were determined in sediments of this ecosystem for the first time. The variation of sediment environmental quality with the sampling period and the differentiation of samples in three groups along the bay were obtained. The usefulness of the multivariate statistical techniques employed for the environmental interpretation of a limited dataset was confirmed.     

Analysis of spatial and temporal water pollution patterns in Lake Dianchi using multivariate statistical methods

Various multivariate statistical methods including cluster analysis (CA), discriminant analysis (DA), factor analysis (FA), and principal component analysis (PCA) were used to explain the spatial and temporal patterns of surface water pollution in Lake Dianchi. The dataset, obtained during the period 2003–2007 from the Kunming Environmental Monitoring Center, consisted of 12 variables surveyed monthly at eight sites. The CA grouped the 12 months into two groups, August–September and the remainder, and divided the lake into two regions based on their different physicochemical properties and pollution levels. The DA showed the best results for data reduction and pattern recognition in both temporal and spatial analysis. It calculated four parameters (TEMP, pH, COD_Mn, and Chl-a) to 85.4% correct assignment in the temporal analysis and three parameters (BOD, NH 2009/11/20Heavy metals (Cd, Cu, Fe, Mn and Zn) concentrations were determined in different tissues (muscle, kidney, liver, brain, gonads, heart and feathers) of Glaucous Gulls (Larus hyperboreus) from Bjørnøya and Jan Mayen. The age and spatial dependent variations in heavy metals were quantified and interpreted in view of the three chemometric techniques, i.e. non-parametric Mann–Whitney U test, redundancy gradient analysis and detrended correspondence analysis. The Glaucous Gulls from Bjørnøya contained significantly higher (p?<?0.05) levels of Cd, Cu and Zn than those inhabited Jan Mayen. Adult birds were characterized by greater (p?<?0.01) concentration of muscle, hepatic and renal heavy metals in comparison to chicks. Insignificantly higher slope constant Zn/Cd for the liver than for the kidney may reflect insignificant Cd exposure. Estimate of transfer factor (TF) allows us to assess variations in heavy metal concentrations during the individual development of Glaucous Gulls. It may be stated that there is a distinct increase of bioaccumulation of all the studied metals during subsequent stages of the bird life.     

A new four-step hierarchy method for combined assessment of groundwater quality and pollution

A new four-step hierarchy method was constructed and applied to evaluate the groundwater quality and pollution of the Dagujia River Basin. The assessment index system is divided into four types: field test indices, common inorganic chemical indices, inorganic toxicology indices, and trace organic indices. Background values of common inorganic chemical indices and inorganic toxicology indices were estimated with the cumulative-probability curve method, and the results showed that the background values of Mg²⁺ (51.1 mg L^?1), total hardness (TH) (509.4 mg L^?1), and NO₃ ^? (182.4 mg L^?1) are all higher than the corresponding grade III values of Quality Standard for Groundwater, indicating that they were poor indicators and therefore were not included in the groundwater quality assessment. The quality assessment results displayed that the field test indices were mainly classified as grade II, accounting for 60.87% of wells sampled. The indices of common inorganic chemical and inorganic toxicology were both mostly in the range of grade III, whereas the trace organic indices were predominantly classified as grade I. The variabilities and excess ratios of the indices were also calculated and evaluated. Spatial distributions showed that the groundwater with poor quality indices was mainly located in the northeast of the basin, which was well-connected with seawater intrusion. Additionally, the pollution assessment revealed that groundwater in well 44 was classified as “moderately polluted,” wells 5 and 8 were “lightly polluted,” and other wells were classified as “unpolluted.”     

On the use of multivariate statistical methods for combining in-stream monitoring data and spatial analysis to characterize water quality conditions in the White River Basin,Indiana, USA

Mechanistic hydrologic and water quality models provide useful alternatives for estimating water quality in unmonitored streams. However, developing these elaborate models for large watersheds can be time-consuming and expensive, in addition to challenges that arise during calibration when there is limited spatial and/or temporal monitored in-stream water quality data. The main objective of this research was to investigate different approaches for developing multivariate analysis models as alternative methods for rapidly assessing relationships between spatio-temporal physical attributes of the watershed and water quality conditions in monitored streams, and then using the developed relationships for estimating water quality conditions in unmonitored streams. The study compares the use of various statistical estimates (mean, geometric mean, trimmed mean, and median) of monitored water quality variables to represent annual and seasonal water quality conditions. The relationship between these estimates and the spatial data is then modeled via linear and non-linear multivariate methods. Overall, the non-linear techniques for classification outperformed the linear techniques with an average cross-validation accuracy of 79.7%. Additionally, the geometric mean based models outperformed models based on other statistical indicators with an average cross-validation accuracy of 80.2%. Dividing the data into annual and quarterly datasets also offered important insights into the behavior of certain water quality variables impacted by seasonal variations. The research provides useful guidance on the use and interpretation of the various statistical estimates and statistical models for multivariate water quality analyses.     

Assessment of water quality using multivariate statistical techniques in the coastal region of Visakhapatnam,India

The present study was intended to develop a Water Quality Index (WQI) for the coastal water of Visakhapatnam, India from multiple measured water quality parameters using different multivariate statistical techniques. Cluster analysis was used to classify the data set into three major groups based on similar water quality characteristics. Discriminant analysis was used to generate a discriminant function for developing a WQI. Discriminant analysis gave the best result for analyzing the seasonal variation of water quality. It helped in data reduction and found the most discriminant parameters responsible for seasonal variation of water quality. Coastal water was classified into good, average, and poor quality considering WQI and the nutrient load. The predictive capacity of WQI was proved with random samples taken from coastal areas. High concentration of ammonia in surface water during winter was attributed to nitrogen fixation by the phytoplankton bloom which resulted due to East India Coastal Current. This study brings out the fact that water quality in the coastal region not only depends on the discharge from different pollution sources but also on the presence of different current patterns. It also illustrates the usefulness of WQI for analyzing the complex nutrient data for assessing the coastal water and identifying different pollution sources, considering reasons for seasonal variation of water quality.