INTEGRATED HYDROLOGIC MODEL CALIBRATION UNDER NON-STATIONARY CLIMATES

With global climate change, water management under non-stationary climate conditions will become a major challenge for scientists. Management decisions are often made based on hydrologic model predictions, yet rarely are these models calibrated under the climate conditions projected in the future. The objective of this work is to investigate the importance of including extreme climatic events, which may be more frequent in the future, in hydrologic model calibration datasets. In this study, the integrated hydrological model HydroGeoSphere (HGS) is applied to the Harold L. Disney Training Center (HLDTC) site in Kentucky, USA, using observed precipitation data and hydrologic observations. The model is calibrated using PEST with different subsets of measured groundwater levels that include and exclude extreme precipitation events (floods) that occurred during the observation period. The results indicate that the inclusion of extreme precipitation events changes the calibrated hydraulic conductivity in all zones, while the largest changes occurred on surface. This work is to aid in future water resource management planning by further informing the calibration process to develop more robust, representative models.