Urbanization and industrialization significantly impact water quality, and detecting the specific factors which influence water quality change would greatly improve urban water environment management. In this study, an improved entropy-weighted matter-element method is used to assess the variations of water quality in two regions with different levels of urbanization in the Yangtze River Delta. Redundancy analysis was used to detect the effects of different industries on water quality. Results show that (1) an improved entropy weight-based matter-element method measures weights of pollutants and water quality levels more reliably and accurately; (2) the improvement rate of water quality in highly urbanized regions is 42.9% during 2005-2014 which is 17.2% higher than that in regions with low urbanization; (3) a decreasing concentration of total phosphorus is the main reason for changes of water quality in both regions, with decreasing concentrations of permanganate index and ammonium nitrogen having a strong influence on changes of water quality in the highly urbanized regions; (4) the decreasing proportion of fishery and heavy industries and the increasing proportion of the tertiary industries significantly influence water quality in highly urbanized regions while the decreasing proportion of animal husbandry is the most important factor influencing the changes of water quality in lowly urbanized regions.
Keywords: Improved entropy weight-based matter-element method; Industry structures; Redundancy analysis; The Yangtze River Delta; Urbanization; Water quality.