Groundwater quality index based on PCA: Wadi El-Natrun, Egypt

Abstract

Groundwater is one of the water's main sources for domestic, agricultural, and industrial uses in arid and semi-arid regions such as the Egyptian western desert. Extensive extraction of the groundwater may lead to the extraordinary decline of the groundwater level coupled with salinization and quality deterioration. This paper aims to investigate the quality of drinking and irrigation water in 47 pumping wells distributed in Wadi El-Natrun, Egypt. Twenty-three hydrochemical parameters that reflected the complexity of the water quality were considered and evaluated. Hierarchical Clustering Analysis (HCA) and Principal Component Analysis (PCA) were sequentially applied to produce potential clusters/groups (groundwater quality classes), classify the groundwater quality data into meaningful classes, and identify the most critical parameters in the classification. HCA produces four major clusters. Electrical conductivity has a high correlation with the Total Dissolved solids “TDS”, Sodium, Calcium, Magnesium, Sulfate, and Strontium. PCA deals with highly correlated parameters and reduces them to a few uncorrelated principal components so that the dimensionality of the transformed data is reduced. Hence, the first three principal components were used to group groundwater quality parameters in comparison to HCA. The grouping provided by the HCA strongly reflects the effect of the first three principal components, showing that the two analyses gave comparable results.In fact, the study proposes a modified Ground Water Quality Index (GWQI) based on the weighted GWQI developed by Tiwari and Mishra (1985). The weights of the studied parameters were estimated based on the PCA where only seven PCs covering about 80.5% of total variance, were used. It has been founded that, the resulting weights match well with the classification performed by the two described methods. The proposed method was used to evaluate the suitability of the water for drinking and agricultural uses based on both WHO and FAO standards. Among the 47 studied wells, only five wells could be considered suitable and good for drinking and 18 wells for irrigation. Also, the water samples are characterized by a high concentration of Sodium, Sulfate, Chloride, and Strontium in addition to TDS.