Modeling Sapanca lake with one and two layer pamolare model for evaluation of water quality

Abstract

Lake Sapanca is one of the major water resources supplying drinking water of northwestern Turkey. The determination of water quality of drinkable water resources is so crucial. There are lots of methods to estimate the water quality. In recent years, mathematical models become so popular for investigation of the water quality. In this research, limiting element which is so important for eutrophication is determined by using nitrogen and phosphorus ratio. Beside that, Vollenweider method and probabilistic approaches used to determine trophic character of Lake Sapanca. To see the relation between phosphorus and chlorophyll-a non linear regression is applied. The thermal stratficiation is investigated by using temperature profile of lake then epilimnion and hypolimnion depth is determined. Lastly, dynamic model named as PAMOLARE used for evaluation water quality of Sapanca Lake. PAMOLARE is a model which is estimating future characteristics of lake by using past data and input loads. Two types models were used; namely: 1-Layer PAMOLARE model and 2-Layer PAMOLARE model. 1-Layer Model assumes that lake is well mixed and no stratification occurs in lake. 1-Layer Model makes estimation by using lake morphology and assuming a constant nutrient loading. 2- Layer Model is more complex model. This model separates lake to two layers as hypolimnion and epilimnion. The mathematical formulas are applied differently to two layers. Beside that 2-Layer Model uses daily environmental data and estimates future data by using these past environmental data. In the construction process of model, general behavior of lake was described. Missing data was obtained by curve fitting programs. Solar intensity, epilimnion and hypolimnion layers depth were described by the physical properties of lake. And different results are obtained according to different scenarios. As a conclusion, water quality of Sapanca Lake is investigated and the PAMOLARE models have been successfully applied to determine future in-lake parameters.