A feasible method to assess inaccuracy caused by patchiness in water quality monitoring

Abstract

This research explains the background processes responsible for the spatial distribution of hydrochemical properties of the picturesque eutrophic Himalayan Lake, Dal, located in Kashmir valley, India. Univariate and multivariate statistical analyses were used to understand the spatiotemporal variability of 18 hydrochemical parameters comprising of 12,960 observations collected from 30 sampling sites well distributed within the lake at a grid spacing of 1 km2 from March 2014 to February 2016. Hierarchical cluster analysis (HCA) grouped all the sampled data into three clusters based on the hydrochemical similarities, Discriminant analysis also revealed the same clusters and patterns in the data, validating the results of HCA. Wilk’s λ quotient distribution revealed the contribution of ions, nutrients, secchi disk transparency, dissolved oxygen and pH in the formation of clusters. The results are in consonance with the Principal Component Analysis of the whole lake data and individual clusters, which showed that the variance is maximally explained by the ionic component (46.82%) followed by dissolved oxygen and pH (9.36%), nitrates and phosphates (7.33%) and Secchi disk transparency (5.98%). Spatial variability of the hydrochemistry of the lake is due to the variations in water depth, lake water dynamics, flushing rate of water, organic matter decomposition, and anthropogenic pressures within and around the Dal lake ecosystem. Overall, the water quality of the lake is unfit for drinking due to the presence of coliform bacteria in the lake waters.