Sub-daily precipitation-streamflow modelling of the karst-dominated basin using an improved grid-based distributed Xinanjiang hydrological model

Abstract

Study region: Hekou karst basin, southwestern China. Study focus: The duality and complexity of hydrological behavior of karst systems pose challenges in precipitation-streamflow modelling, reliable hydrological models with high temporal resolution are warranted for predicting floods in humid flood-prone karst-dominated basins. This study proposes a grid-based distributed karst Xin’anjiang hydrological model (DK-XAJ), the model structure of each grid was developed using the conceptual Xin’anjiang model. Water exchange among each grid and the spatial heterogeneities of land use types and the karstification in each grid were considered. Information on two new linear reservoirs was added to represent the rapid-conduit and slow-matrix underground flow to improve the karst dual-porosities, the range of the underground runoff partitioning parameter was estimated from the observed flood recessions. New hydrological insights: The DK-XAJ model demonstrated good successive hourly streamflow modelling performance with mean values of 0.84, 0.83, 0.28, and 0.88 for NSE, KGE, RRE, and R2, and results during validation of one nested interior grid without recalibration was also good. The results indicate that the DK-XAJ model can help obtain detailed results for successive streamflow processes, the peaks of floods under different magnitudes and the marked increase and recession of the flash floods can be accurately reproduced and depicted. The DK-XAJ model can therefore be considered a new tool for the prediction of sub-daily precipitation-streamflow and flood events in karst-dominated basins.

Keywords

• Karst hydrology
• Xinanjiang hydrological model
• Distributed hydrological model
• Hourly successive streamflow simulation