GPM-based Conversion of Daily to Hourly Rainfall Data for Flood Modelling in Kuranji

Abstract

Tapin District, located in the Kuranji Watershed of South Kalimantan Province, frequently experiences flooding due to heavy rainfall. Flood event simulation can be utilized to generate hourly discharge data for flood early warning systems, addressing the absence of observed hourly discharge data from the Kuranji Automatic Water Level Recorder (AWLR). Moreover, limited watershed parameters and the unavailability of hourly rainfall data pose challenges in developing a hydrological model for the Kuranji Catchment. To overcome this issue, the hourly rainfall distribution pattern from 10 Global Precipitation Measurement Mission (GPM) Satellite grids were used to convert daily rainfall observations recorded at the Talaga Langsat, Lok Paikat, and Tapin Utara rainfall stations Hourly rainfall data from each GPM grid were matched with several synthetic hourly rainfall distribution patterns. The chosen distribution pattern was subsequently applied to simulate flood events using the HEC-HMS model. Two flood events, dated 4–7 March 2017 and 21–23 March 2018, were used for trial fitting to estimate catchment parameters. Simulated hourly discharges were compared with daily discharge data from the AWLR at Kuranji station, using Total Relative Volume Error (TRVE) and Root Mean Square Error (RMSE) as performance indicators. The simulation results demonstrate that daily-to-hourly rainfall conversion is applicable for estimating catchment parameters, with average TRVE and RMSE values of 1.955% and 7.025 m³/s, respectively. Furthermore, the trial fitting results indicate that the simulated discharge values align well with observed peak daily discharge data. Acquiring and incorporating hourly discharge data would enhance the accuracy of daily-to-hourly rainfall conversion by synchronizing the temporal resolution. Additionally, incorporating more flood events into trial fitting tests could produce more robust and representative outcomes.