Investigating river water quality assessment through non‐parametric analysis: A case study of the Godavari River in India

Abstract

AbstractThis study examines temporal variations in water quality parameters of Godavari River. The present study examines the concentrations of seven water quality parameters, namely total alkalinity (Alk), calcium (Ca), chlorine (Cl), total hardness (Hard), magnesium (Mg), sodium (Na), and sulfate (SO4), using data from 1981 to 2005 at seven gauging stations, that is, Mancherial, Pathagudem, Nowrangpur, Jagdalpur, Perur, Konta, and Polavaram. A seasonal analysis (pre‐monsoon, monsoon, and post‐monsoon) is conducted to detect trends in water quality parameters. The study also uses normality tests Kolmogorov–Smirnov (K–S) and Shapiro–Wilk (S–W) which reveals nonnormal distribution in the majority of datasets, necessitating non‐parametric analysis. The concentration of Alk, Ca, and Hard is found to be significantly increasing trend at some stations during all three seasons, along with significantly decreasing trend found in the concentrations of Cl, Na, and SO4 using non‐parametric Mann–Kendall's (M–K) and Sen's Slope test. The concentration of Mg depicts mixed trends at different stations which may be due the variation in agricultural practices or industrial activities surrounding the stations. It is evident that none of the parameter exhibits statistically significant trends across all seasons at any station. The results of Sen's Slope test are found to be in good agreement with Mann–Kendall's test statistics. A second order auto‐regressive model (AR (2)) is successfully employed to predict and forecast the water quality parameters in present and future time steps. The authenticity of AR (2) model is assessed using Nash‐Sutcliffe (NSE) efficiency and coefficient of determination (R2). The values of NSE and R2 are found to be more than 0.85 for all parameters during all the three seasons, along with, the values of mean square error (MSE), root mean square error (RMSE), and mean absolute error (MAE). The results of coefficient of variation (CV) and persistence‐criterion (PC) also depicts that AR (2) can successfully be utilized in forecasting the water quality parameters. The research contributes insights for effective water resource management in Godavari River basin.