Author(s): Theerapol Charoensuk; Jakob Luchner; Nicola Balbarini; Piyamarn Sisomphon; Peter Bauer-Gottwein

Linked Author(s):

Keywords: Flood; Forecasting; Performance; Uncertainty; Chao Phraya River basin

Abstract: Accurate predictions of flooding are important for emergency response and flood management. The quality of numerical weather forecasts is one main parameter controlling the quality of flood forecasts. This study focused on improving the skill of the Chao Phraya hydrologic forecasting system using pre-processed numerical weather forecasts. The applied preprocessing method was quantile mapping bias correction (QM). The rainfall forecast data with QM (WRF-QM) pre-processing were evaluated against the daily rainfall measurements from Hydro- Informatics Institute (HII) 's stations in each sub-catchment and compared with the raw rainfall prediction (Raw WRF). Evaluation results for the period 2016 – 2019 (training period) and 2020-2021 (testing period) showed that the WRF-QM method effectively improves the rainfall prediction by around 15 percent. Reforecasting experiments were performed with the hydrological model using pre-processed forecast precipitation to compare runoff forecast with and without the pre-processing method. The overall performance results show that runoff forecasting with WRF-QM preprocessing can significantly improve accuracy by ca. 30 percent when compared to runoff without preprocessing (Raw WRF). Our study demonstrates that the pre-processed numerical weather forecasts can improve the skill of hydrologic forecasting systems to benefit the real-time Chao Phaya flood operation system.

DOI: https://doi.org/10.3850/978-90-833476-1-5_iahr40wc-p0343-cd

Year: 2023