Predicting Typhoon Morakot's Catastrophic Rainfall and Flooding With a Cloud-Scale Ensemble System

This study examines the prediction and predictability of the recent catastrophic rainfall and flooding event over Taiwan induced by Typhoon Morakot (2009) with state-of-the-art numerical weather prediction models. It is found that a cloud-resolving mesoscale ensemble, initialized with analysis and flow-dependent perturbations obtained from a realtime global ensemble data assimilation system, were able to predict this record-breaking flooding event, producing probability forecasts potentially valuable to the emergency management decision makers and the general public. Since all the advanced modeling and data assimilation techniques used here are readily available for realtime operational implementation provided sufficient computing resources are made available, this study demonstrates the potential and need of using ensemble-based analysis and forecasting, along with enhanced computing, in predicting extreme weather events like Typhhon Morakot at operational centers.