Nitrate processes investigation for improved ecohydrological modeling at the catchment scale
The excessive presence of nitrate in predominantly agricultural catchments can result in serious negative impacts on water resources and ecosystems. Ecohydrological models can be used to investigate nitrate dynamics in water and soil and future activities for water quality improvement. The Treene catchment (northern Germany), is an appropriate study region since is predominantly agricultural and presents significant water quality problems caused by nitrate. This thesis contributes to this subject by investigating the representation of nitrate processes in models through the analysis of temporal parameter sensitivity. This investigation identifies the dominant model parameters, and processes, for each time step. The dominant temporally sensitive parameters are then used for calibrating a model. Thus, a new calibration procedure with distinct performance measures is developed to represent both discharge and nitrate loads accurately with the same parameter set. Furthermore, different Best Management Practices (BMPs) are implemented in the model to assess their efficiency in reducing nitrate loads at the catchment scale. The dominant temporal nitrate parameters identified in the first step are related to nitrate transport with runoff components and to plant uptake processes. The calibration procedure resulted in a more reliable model run in all discharge and nitrate load phases, since it considers dynamics and magnitudes of discharge and nitrate loads in a balanced way. The simulation of BMPs indicated higher nitrate reductions by using BMPs containing fertilization reduction and buffer strips. The BMPs simulation demonstrated the complexity of decision making facing possibly contrasting regulations. The methods proposed and tested contribute to improve the representation of nitrate processes and their dynamics in models to obtain reliable simulation results to address ecohydrological challenges at the catchment scale.