Improve the Xinanjiang model based on various hydrological model structures

Abstract

【Objective】 The Xinanjiang model is a hydrological model widely used for catchment modelling, but its accuracy depends on many factors. The objective of this paper is to present ways to improve the Xinanjiang model. 【Method】 The study is based on the Wangjiaba sub-catchment in the Huaihe River basin. Three models, including the conventional Xinanjiang model, the GR4J (mode′le du Ge′nie Rural a′4 parame′tres Journalier) model and the HYMOD model, were used to model flow in the sub-catchment. Based on the convolution structure of the GR4J and the HYMOD model, we proposed two methods to improve Xinanjiang model. 【Result】 ①All three models simulated the flow in the sub-catchment well, but the Xinanjiang model was the best, despite its underestimation of the flow when discharge was large. ② The simulation results using the Xinanjiang model combined with the confluence structure of the HYMOD model were comparable to those simulated using the Xinanjiang model. ③ Combining the Xinanjiang model with the confluence structure of the GR4J model improved the simulation accuracy when discharge flow was large. 【Conclusion】 Combining the Xinganjiang model with the confluence structure of the GR4J model can resolve the underestimated runoff at peak discharge by the Xinganjiang model. The proposed model can hence improve catchment modelling.

Keywords

• comparison of hydrological models
• xinanjiang model
• gr4j model
• hymod model
• runoff simulation