Modeling contaminant transport in the Marmara region and analyzing the associated parametric sensitivity

Abstract

The increasing human activities, urbanization and land use changes in recent years have contributed to water quality degradation by affecting the transport of contaminants to receiving waters by overland flow. This necessitates the development of modeling approaches for planning and management of catchments that play a significant role on water supply. Geographical Information Systems (GIS) has become an important methodological approach in catchment modeling with the facilities to obtain spatial data and advanced visualization of topographical features. The main objective of this study is to determine the effects of land use changes on nutrient and heavy metal loads by modeling the transport of these contaminants in the Marmara Region and to assess the sensitivity of parameters affecting the transport processes. Storm Water Management Model (SWMM) was chosen to develop the catchment hydrological model. GIS based maps and data including land use and cover were used to determine catchment boundaries and develop the input parameters required for SWMM with time series rainfall data. Total nitrogen, total phosphorus, copper, zinc and nickel concentrations in runoff were estimated by using previously analyzed soil samples collected from the different lands of the Marmara Region. Spatial data and concentrations obtained were transferred to SWMM to develop the catchment hydrological model and describe the relationships between contaminant load and land use. Finally, parametric sensitivity analysis was carried out to determine the most important parameters affecting SWMM model outcomes. The model results revealed that increase in rainfall intensities and % imperviousness values contribute to runoff production and associated contaminant loads in catchments. Area of subcatchments, rainfall and maximum buildup of contaminants are the most significant parameters in SWMM to predict runoff and event loads. Percent imperviousness and percent slope are the least significant parameters amongst other parameters affecting output.