Comprehensive Water Quality Indexing Using Functional Data and Principal Component Analysis for River Narmada

Abstract

The degrading water quality status of River Narmada can be deciphered by a tool called Water Quality Indexing.From the entire river basin seven important sites are chosen.Water Quality pertaining to twenty-five parameters are analyzed from 1991-2013.The final results are averaged to get the indices for a typical year.The current study commences with some preliminary methods.A significance test is conducted to determine three relevant parameters from each site. Based on these relevant parameters,Water Quality Indexing is done using Functional Data Analysis and Principal Component Analysis and finally error analysis of the methods is done.The preliminary analysis show that Biochemical Oxygen Demand,Dissolved Oxygen,Chemical Oxygen Demand and Total and Faecal Coliform are the most crucial water quality variables along the mainstream of Narmada River Basin.Both Principal Component Analysis and Factor Analysis yield seven components from which it is concluded that agricultural runoff and industrial wastes account for an increase in alkalinity in the river water.It is also concluded that the deterioration in Dissolved Oxygen is being rejuvenated by variables causing alkalinity in water. Conventional Water Quality Indexing done along the seven sites shows that for Station 1,2,5,6 and 7,water quality remains excellent almost throughout the year except for Stations 2 and 4.Functional Data Analysis and Principal Component Analysis also shows similar results.From the error analysis Functional Data Analysis using Entropy Weights is found to have least error.Thus for a typical year the minimum indices implying the worst water quality are shown in a map which can help in deciphering the water quality status of the river well.