XXT is a newly developed semi-distributed rainfall-runoff mode based on the soil moisture storage capacity distribution curve, the highlight of the Xinanjiang model, together with the simple model structure of TOPMODEL. It performs better than the traditional hydrological models TOPMODEL and Xinanjiang for daily runoff and flood simulations over the various watersheds in different size and climatic dimensions in China. However, XXT performs worse in daily
stream flow simulating compared to ANN-based rainfall-runoff models, especially for large-scale basins. The objective of the present study is therefore to enhance XXT performance in daily stream flow modelling by integrating artificial neural network (ANN) into it. A new hybrid model entitled as ANN-XXT is proposed in this work. Yingluoxia watershed (10009km2), situated in the arid, semi-arid region of northwestern China, in the upper stream of the Heihe River Basin, is selected as a large-scale basin for testing the new model. The results show that the daily stream flows
predicted by the new model are in very good agreement with the observed ones, while those simulated by XXT underestimate the main peak-flows and are distinct from the observed daily stream flow for low flow stages for both the calibration and validation period. The results indicate that the proposed integrated model based on ANN and XXT has promise in daily stream flow simulating for large-scale basins.