Integrated water system simulation by considering hydrological and biogeochemical processes: model development, with parameter sensitivity and autocalibration
- 1Key Laboratory of Water Cycle and Related Land Surface Processes, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, 100101, China
- 2CSIRO Digital Productivity Flagship, Leeuwin Centre, 65 Brockway Road, Floreat Park, WA 6014, Australia
- 3College of Global Change and Earth System Science, Beijing Normal University, Beijing, 100875, China
- 4CSIRO Agriculture Flagship, GPO BOX 1666, Canberra, ACT 2601, Australia
Abstract. Integrated water system modeling is a feasible approach to understanding severe water crises in the world and promoting the implementation of integrated river basin management. In this study, a classic hydrological model (the time variant gain model: TVGM) was extended to an integrated water system model by coupling multiple water-related processes in hydrology, biogeochemistry, water quality, and ecology, and considering the interference of human activities. A parameter analysis tool, which included sensitivity analysis, autocalibration and model performance evaluation, was developed to improve modeling efficiency. To demonstrate the model performances, the Shaying River catchment, which is the largest highly regulated and heavily polluted tributary of the Huai River basin in China, was selected as the case study area. The model performances were evaluated on the key water-related components including runoff, water quality, diffuse pollution load (or nonpoint sources) and crop yield. Results showed that our proposed model simulated most components reasonably well. The simulated daily runoff at most regulated and less-regulated stations matched well with the observations. The average correlation coefficient and Nash–Sutcliffe efficiency were 0.85 and 0.70, respectively. Both the simulated low and high flows at most stations were improved when the dam regulation was considered. The daily ammonium–nitrogen (NH4–N) concentration was also well captured with the average correlation coefficient of 0.67. Furthermore, the diffuse source load of NH4–N and the corn yield were reasonably simulated at the administrative region scale. This integrated water system model is expected to improve the simulation performances with extension to more model functionalities, and to provide a scientific basis for the implementation in integrated river basin managements.