Fuzzy estimation of trophic state using satellite data

Abstract

Eutrophication is one of the major water quality problems in developing and developed countries. Generally, trophic state of water sources is under a strict watch and is determined by in situ water sampling for some important water resources. However, a limited number of water samples provide insufficient statistical confidence with an overall eutrophic status by using point-basis water sampling data which is usually executed under a physical and financial limitation. Also, traditional trophic state indices, such as Carlson index and OECD index, employ a crisp determination that often causes debated argument. This research applies fuzzy set theory on satellite data to describe the trophic state for reservoir water. A multi-spectral SPOT satellite image was converted into water quality variables, such as phosphorus, Secchi depth, and chlorophyll that are the major affecting factors of eutrophication. Fuzzy synthetic evaluation for the trophic state was developed by using satellite-derived two-dimensional water variables. Feitsui Reservoir, which is the most important water supply for over five million people in the great Taipei area, Taiwan, was the study site for demonstrating the trophic state determination through the fuzzy evaluation method.