Development of an R-CLIPER model using GSMaP and TRMM precipitation data for tropical cyclones affecting Vietnam

Tropical cyclone (TC)-induced rainfall is one of the most extreme rain phenomena, usually resulting in severe flooding and inundation when a TC makes landfall and is still considered the biggest challenge for TC forecasting. This study focuses on developing a R-CLIPER (Tropical Cyclone Rainfall Climatology and Persistence) model for TC-induced rainfall forecasting over the Vietnam region using precipitation data from GSMaP (Global Satellite Mapping of Precipitation) and TRMM (The Tropical Rainfall Measuring Mission). We used the best track data for 199 TCs affecting the Vietnam region during the period 2000–2021, obtained from the Regional Specialized Meteorological Center (RSMC) Tokyo Typhoon Center. The performance of the established R-CLIPER models for TC rainfall forecasting was assessed under different circumstances by categorizing the rain thresholds and TC intensities. The evaluations are conducted by comparing the 24-hour R-CLIPER model’s rainfall forecast against satellite-estimated and surface-observed rains, indicating a reasonable prediction of two model versions, namely, the R-GSMaP and R-TRMM models, with slight outperformance of the R-GSMaP over the R-TRMM model. In detailed analyses of three case studies, we found that the results of predicted TC-induced rainfall largely vary depending on both weather conditions and TC tracks. In addition, using the developed R-CLIPER model as a baseline, we evaluate the performances of three global NWP models in TC-induced rainfall prediction by calculating a percentage of improvement in the statistical scores over those of the R-CLIPER model. The results revealed the greatest improvement in the forecast by the Integrated Forecasting System (IFS) model.