Modeling Impreciseness of Trophic State Levels for Eutrophication Assessment

Abstract

Tremendous efforts have been madeto propose numerous indices for quantifyingstates of eutrophication in lentic water bodies over the last decades. Most of the indices are developed or modified considering conventional trophic state indices (e.g., Carlson trophic state index (TSI)) or pre-specified ranges of cause and response variables (e.g., index proposed by Organization of Economic and Cooperation and Development (OECD)), where a strong correlation structure must exist among cause and response variables. Here an attempt is made to model impreciseness or subjectivity of trophic state levels,especially for weakly correlated variables. A generalized framework using fuzzy synthetic evaluation (FSE), analytic hierarchy process (AHP) and the concept of statistical entropyis proposed to assess the trophic state of lentic water bodies. The proposed methodology has been tested with the case study of Jordan lake data set collected from the United States of Geological Survey (USGS) data base.The result shows that, the eutrophication problem is critical mostly during the months of August to October,and the method can efficiently capture the rapid changes in trophic states.The sensitivity of the model has been tested with different weight structures for cause and response variables considered during model development, and the robustness of the model has also been tested with different weighting techniques.