Analysis and Simulation of Human Activity Impact on Streamflow in the Huaihe River Basin with a Large-Scale Hydrologic Model

A hydrologic model coupled with a land surface model is applied to simulate the hydrologic processes in the Huaihe River basin, China. Parameters of the land surface model are interpolated from global soil and vegetation datasets. The characteristics of the basin are derived from digital elevation models (DEMs) and a national geological survey atlas using newly developed algorithms. The NCEP-NCAR reanalysis dataset and observed precipitation data are used as meteorological inputs for simulating the hydrologic processes in the basin. The coupled model is first calibrated and validated by using observed streamflow over the period of 1980-87. A long-term continuous simulation is then carried out for 1980-2003, forced with observed rainfall data. Results show that the model behavior is reasonable for flood years, whereas streamflows are sometimes overestimated for dry years since the 1990s when water withdrawal increased substantially because of the growing industrial activities and the development of water projects. Observed streamflow and water withdrawal data showed that human activities have obviously affected the surface rainfall-runoff process, especially in dry years. Two methods are proposed to study the human dimension in the hydrologic cycle. One method is to reconstruct the natural streamflow series using local volumes of withdrawals. The simulated results are more consistent with the reconstructed hydrograph than the initially observed hydrograph. The other method is to integrate a designated module into the coupled model system to represent the effect of human activities. This method can significantly improve the model performance in terms of streamflow simulation.