Đánh giá sản phẩm mưa dự báo tổ hợp từ các mô hình dự báo mưa số trị toàn cầu: ứng dụng cho lưu vực sông Kone
Abstract
The use of rainfall forecasts derived from numerical weather prediction (NWP) models has recently become a common approach to improve the lead time of streamflow forecasts for flood control and real-time reservoir operation. However, evaluation of rainfall forecasts is often required before it can be used as input to hydrological models for flood forecasting. This paper, therefore, presents an assessment of NWP-based ensemble forecasts of rainfall for flood forecasting. Four global NWP models are chosen for this study including ECMWF, JMA, KMA, and NCEP. The assessment is carried out for the Kone river basin in Vietnam during flood seasons (September to December) during 2014–2018. The results are compared using multiple deterministic and probabilistic forecast verifications calculated at the original model grid value of 0.5o resolution, compared to gridded rainfall station data at different lead times (up to +10 days). The results show that ECMWF has the highest skill in rainfall forecast over the basin while NCEP showed the worst skillful rainfall forecasts in most cases considered. The findings of the paper have important implications for the use of ensemble NWP rainfall forecasts in hydrological modeling, particularly for flood forecasting.
Keywords: ensemble rainfall forecast, numerical weather prediction model, NWP-based ensemble forecasts, Kone river
