ABSTRACT
Heavy precipitation events recurrently occur during monsoon in Nepal. Also, climate change projections indicate an increasing trend in the frequency of such events. In this study, we firstly applied a kinematic wave flow model on hill slopes for simulating surface runoff. A two-dimensional (2D) shallow-water unsteady flow model was then used to simulate the inundation process. We investigated the extreme precipitation event of 14–15 Aug 2014 over West Rapti River (WRR) basin, Nepal. During the 24 h, precipitation amount of 528 mm was recorded at one climate station (highest ever recorded in Nepal) near to the selected basin. We used ground-based hourly rainfall data and a hydrological model was validated against ground-based hourly water level at Kusum outlet of WRR basin. The simulated surface runoffs from different hill slopes were then inputted for inundation analysis. Our results show similar spatial patterns of inundation with those derived from satellite images. Since it was difficult to get observed inundation depths our study still could not validate quantitatively the inundation. This study also attempts to analyze the effectiveness of satellite-derived rainfall in capturing extreme precipitation events. We used hourly data of 1) PERSIANN (Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks), resolution (25 km x 25 km), and 2) PERSIANN-Cloud Classification System (PERSIANN-CCS), resolution (4km x 4km). In general, satellite-derived rainfall highly underestimates extreme precipitation with respect to ground-based data in the study area. This study highly recommends the inclusion of bias correction of satellite-derived rainfall.
