On the utilization of RDCA method for detecting and predicting the occurrence of heavy rainfall in Indonesia

Abstract

Accurate prediction of heavy rainfall events is crucial to mitigate hydrometeorological disasters. The real-time cloud observation by weather satellites Himawari-8 has enabled researchers to develop a nowcasting method (short-term prediction) to identify rainfall events in advance, namely the Rapidly Developing Cumulus Area (RDCA) method. The RDCA index, which is the product of the RDCA method, is the probability of occurrence of heavy rainfall. It has been evaluated and successfully utilized in Japan. However, the performance of the RDCA in the tropics has not been assessed yet. Indonesia —as a tropical country consisting of abundant islands— exhibits distinct climate and convective systems due to its geographical characteristics. Different temporal resolution of Himawari-8 data between Japan region (2.5 min) and the tropics (10 min) may also result in different RDCA performance. Therefore, it is important to conduct an evaluation study before operating the RDCA in Indonesia. By performing spatial comparison and using dichotomous and probabilistic verifications with rain gauges at various thresholds, this study finds that the RDCA is able to detect and predict rainfall events in Indonesia. The best lead time is 10 min but the predictive signals extend up to 30 min in advance; it is shorter than that observed in Japan. Lighter rainfall has higher predictability compared with heavier rainfall. The predicted heavy rainfall events show good accuracy, probability of detection, and brier score. The RDCA index and rainfall intensity appear to have a good correlation, where light (heavy) rainfall is better predicted by low (high) RDCA indices. It confirms the usefulness of varying RDCA index in contrast to the classic dichotomous RDCA index. This study reveals the positive prospects of RDCA in Indonesia. We also discuss some future challenges and give possible recommendations regarding an operational RDCA system. This study will benefit the development of early warning systems in Indonesia.