Investigating Spatial Heterogeneity of Karst Water Storage Capacity and Nonclosure of Underground Watersheds in Karst Hydrological Simulation

Zeling Ren,Binquan Li,Yang Xiao,Kuang Li
DOI: https://doi.org/10.1002/hyp.70012
IF: 3.784
2024-12-03
Hydrological Processes
Abstract:This study introduces a novel karst hydrological module incorporating a karst water storage capacity distribution curve and an underground nonclosure coefficient. The XAJ‐karst model, developed by integrating this module with the Xinanjiang (XAJ) model, exhibits superior performance in both daily and hourly simulations compared to the standard XAJ model in the Xiangyang watershed. The discussion on the dynamic change of the nonclosure coefficient demonstrates its responsive relationship with antecedent‐precipitation and its mutational nature. Karst landforms interfere with the runoff generation and confluence process, resulting in generally poor hydrological simulation accuracy in karst watersheds. We proposed a new karst hydrological module, which has two cores. One is the karst water storage capacity distribution curve that represents the distribution of runoff generation thresholds in karst areas, and the other is the underground nonclosure coefficient that represents the nonclosure phenomenon of underground watersheds in karst areas. The new module was further coupled with the Xinanjiang rainfall–runoff (XAJ) model to establish a complete hydrological model for karst areas (referred to as XAJ‐karst model). The sensitivity of the XAJ‐karst model parameters was analysed using the Sobol method, and applied to a typical karst watershed in Guizhou Province, China, to test the model performance on daily and hourly time scales. In addition, we also explored the impact of dynamic changes in the nonclosure coefficient of underground watershed area in karst watersheds on model results. Results showed that the average value of Kling–Gupta efficiency (KGE) of the XAJ‐karst model on the daily and hourly time scales was 0.85 and 0.77, respectively. In comparison with the XAJ model, the average KGE value of the XAJ‐karst model on both daily and hourly scales improved by 10.8% and 6.4%, respectively, demonstrating better simulation accuracy. In addition, there is a underground nonclosure phenomenon in the Xiangyang watershed, and the actual area of underground watershed expands abruptly as the antecedent‐precipitation increases to the critical value. Moreover, the water storage and hysteresis effects of the karst landform result in a certain hysteresis in water exchange between the underground watershed and adjacent watersheds.
water resources
What problem does this paper attempt to address?