Managing water quality of a river using an integrated geographically weighted regression technique with fuzzy decision-making model.

Abstract

In the recent times, water quality of most of the rivers in India has been steadily degrading due to increasing numbers of point and non-point sources of pollution. The tremendous increase in population, rapid urbanization, change in irrigation patterns, and unplanned growth of industries without proper enforcement of environmental standards are some of the major causes for poor quality of river water. In addition, unpredictable and scanty rainfall is resulting in uncertain natural stream flow which further leads to uncertainty in assessing and predicting the quality of river water. This paper deals with the assessment of the overall status of water quality of a river by developing a fuzzy-based water quality evaluation system. The quality of water needed for different beneficial uses is based on the value of various parameters. Since the quality attributes of the parameters are fuzzy in nature, they have been described by the linguistic variables. The water quality index of each specific site is then calculated by aggregating the attributes with respect to their degree of importance, which is also expressed in the form of linguistic terms. Finally, a case study of the river Yamuna has been carried out to evaluate the fuzzy comprehensive water quality index (FCWQI). In this study, the FCWQI has been determined only for the use of water for drinking purposes though this model can be applied for other uses as well. The FCWQI developed herein is based on an integrated approach, which clearly describes the overall state of the water quality by a single rational number. Spatial and parametric sensitivity of the FCWQI model of the river basin is also determined using GIS-based geographically weighted regression technique. The methodology suggests a novel way of introducing parametric sensitivity in defining water quality indices used for surface water quality assessment.