Modeling Streamflow Dominated by Snowmelt in an Ungauged Basin in Northwestern China

Abstract

The spring and summer peak flow, which is dominated by snowmelt and rainfall, can result in serious damages if flooding occurs. Further, climate change can influence the timing and intensity of the peak flows. To predict the spring peak flow and reduce the damages, better understanding of the runoff formation mechanism is important. In this study a hydrologic model is developed for an ungauged basin, Kaidu River watershed, located in Xinjiang, China. The hydrologic model is based on system dynamic principles and can simulate the daily runoff. The model captures the snowmelt process and the soil infiltration process influenced by the frozen state of soil. The model is calibrated for three different ranges of flows, i.e., high flow, low flow, and mixed flow. Model validation demonstrated that model is robust for all three ranges of flows in terms of estimating spring peak flow and baseflow. The scatter plots and box plots showed that the model can simulate the hydrological process in the Kaidu River basin well. The statistical measures showed that the performance of the model for high flows and mixed flows was good. For the low flow, the calibration is satisfactory while the validation is unsatisfactory. Better estimation of spring peak flow can result in improved management of flows and reduction in flood damages.