Effect of mountainous rainfall on uncertainty in flood model parameter estimation

Abstract

Explaining the significant variability of rainfall in orographically complex mountainous regions remains a challenging task even for modern raingauge networks. To address this issue, a real-time spatial rainfall field estimation model, called WREPN (WRF Rainfall-Elevation Parameterized Nowcasting), has been developed, incorporating the influence of mountain effect based on ground raingauge networks. In this study, we examined the effect of mountainous rainfall estimates on the uncertainty of flood model parameter estimation. As a comparison, an inverse distance weighting technique was applied to ground raingauge data to estimate the spatial rainfall field. To convert the spatial rainfall fields into flood volumes, we employed the ModClark model, a conceptual rainfall–runoff model with distributed rainfall input. Bayesian theory was applied for parameter estimation to incorporate uncertainty analysis. The ModClark model demonstrated good flood reproducibility regardless of the estimation method for spatial rainfall fields. Parameter estimation results indicated that the WREPN spatial rainfall field, which accounted for the influence of the mountain effect, led to lower curve numbers due to higher estimated rainfall compared to the IDW spatial rainfall field, while the concentration time and storage coefficient showed minimal differences. HIGHLIGHTS The study introduces the WREPN model, which incorporates the influence of mountain effect in rainfall estimation, leading to reduced uncertainty in flood model parameter estimation compared to traditional methods like IDW.; Utilizing the ModClark model for converting spatial rainfall fields into flood volumes, the study demonstrates good flood reproducibility regardless of the spatial rainfall field estimation method used.; By applying Bayesian theory for parameter estimation, the study shows that the WREPN spatial rainfall field results in lower curve numbers and more accurate flood estimations compared to the IDW spatial rainfall field, highlighting the importance of considering mountainous rainfall effects in hydrological modeling.;

Keywords

• modclark model
• mountainous watershed
• orographic rainfall
• spatial rainfall field
• uncertainty analysis