Xây dựng mô hình dự báo mưa số trị cho lưu vực sông Vu Gia Thu Bồn

Abstract

Convetional flood forecast models often rely on the observation or real time simulation (based on rainfall information) of discharge and water level at certain river nodes on the catchment. This method offers a forecast lead time that depends on the size of the drainage area and is relatively short for small and steep catchments. Meanwhile, flood forecasts based on quantitative rainfall estimate from numerical weather prediction (NWP) models demonstrate greater accuracy with longer forecast horizon. Therefore, NWP-derived flood forecast is a promising technology. In this study, Weather Research and Forecasting model (WRF) will be applied to downscale the global rainfall dataset forecasted by Japan's GSM model and generate the forecasts from 1-3 days before the storms occur. Results show that the WRF model established for the Vu Gia-Thu Bon river basin provides reasonable forecasts that are comparable to the observed data