Automatic Calibration of a Hydrological Model Using Multiobjective Particle Swarm Optimization and TOPSIS

The performance of a hydrological model heavily depends on choosing suitable model parameters. A framework for automatic calibration of a hydrological model named the Xinanjiang model with multiobjectives has been presented. In the calibration framework, a MOPSO algorithm was employed to find the non-dominated front in the objective space, and an entropy-based TOPSIS ranking method was used to rank the non-dominated solutions. As an application example, daily rainfall, evaporation and flow discharge data are used to calibrate and verify the Xinanjiang model in Misai catchment with an area of 799 km2. The results show that the MOPSO is efficient and robust to find non-dominated front of the Xinanjiang model. The results also show that the entropy-based TOPSIS provides an impersonal method to calculate objective weights and rank the non-dominated solutions.