WATER QUALITY MAPPING AND ANALYSIS IN A STRETCH OF GANGA RIVER

Abstract

A major barrier to water quality modeling is the lack of an efficient system for water quality monitoring. Traditional water quality sampling is time-consuming, expensive, and can only be taken for small size samples. Also, instant and accurate water quality data cannot always be provided to satisfy the demands of water quality modeling and parameter calibration. So, we need a modern approach which is based on Remote sensing and Geographical information System (GIS) for water quality trends. In this Study a decision making tool have been generated for water quality mapping of Ganga River in parts of Kanpur, Allahabad and Varanasi districts of Uttar Pradesh, India. Water sample have been collected from nine stations and analyzed in the laboratory which provides results of various water quality parameters. Digital numbers (DN values) for four bands Blue, Green, Red and NIR of LANDSAT 8 images have been calculated. Observed DN values on those nine sampling stations of each band along with principal component analysis and band ratios are compared with measured water quality parameters. The water quality parameters includes pH, Biochemical Oxygen Demand (BOD), Chemical Oxygen Demand (COD), Dissolve oxygen (DO), Total Solids ('l's), Total Dissolve Solids (TDS), Total Solids (TS), Ammonical Nitrogen, Fluoride, Chloride, Magnesium, Turbidity, Conductivity and Phosphorus. Using DN values of LANSAT 8 images and in situ measured data of water quality parameters correlation and multiple linear regression models have been generated which is based on most appropriate band combinations having highest multiple correlation coefficient, R2 value. Using these multiple regression models, water quality parameters can be predicted for entire study area. Then, applying simple linear discriminate function to each pixel in study area, grouping of these dependent water quality variables into discrete classes achieved the classification to produce water quality maps.