Abstract ID: 149
Subseasonal tropical cyclone precipitation prediction in GEOS-S2S and the WMO S2S models
Lead Author: Chia-Ying Lee
Lamont-Doherty Earth Observatory, Columbia University, United States of America
Keywords: tropical cyclone precipitation, tropical cyclone, Skill analysis, GEOS-S2S, WMO S2S
Abstract: Tropical cyclone precipitation (TCP) is a significant contributor to total annual rainfall and often causes extreme precipitation events in tropical and subtropical regions. In this presentation, we will evaluate the representation of TCP, total rainfall, and their ratio in re-forecasts from the Goddard Earth Observing System – Subseasonal to Seasonal version 2 (GEOS-S2S) prediction system and the models that participate in the World Meteorological Organization (WMO)’s S2S project. Our results show that, for all lead times considered (weeks 1-4) the GEOS-S2S model and the WMO S2S models have dry biases in TCP in the North Atlantic and wet biases in the Eastern North and Western North Pacific basins, when compared to a satellite precipitation product (CMORPH). In terms of the pattern correlation with observations, GEOS-S2S performs better (0.9), compared to the WMO S2S models (0.45 – 0.84). When the biases in TCP are attributed to biases in total rainfall, tropical cyclone activity, and storm-scale precipitation, the frequency biases are found to be the dominant contribution to TCP biases but in some models the biases in storm-scale precipitation contribute significantly to TCP biases at equatorial latitudes. We will then examine the models’ skill in predicting TCP at subseasonal timescales using performance metrics such as the ranked probability skill score and the structure-amplitude-location spatial verification for TCP, and the Brier Skill score for tropical cyclone genesis and occurrence. Our initial results suggest that the prediction skill of TCP in GEOS-S2S and WMO S2S models is comparable. We will further discuss the impact of the TCP climatology representation on prediction skill and its relevance to mean and extreme precipitation prediction on subseasonal timescales.
Co-authors:
Jorge L. García-Franco (Lamont-Doherty Earth Observatory, Columbia University)
Suzana J. Camargo (Lamont-Doherty Earth Observatory, Columbia University)
Michael K. Tippett (Department of Applied Physics and Applied Mathematics, Columbia University)
Andrea Molod (Goddard Space Flight Center, NASA)
Young-Kwon Lim (Goddard Space Flight Center, NASA)
Daehyun Kim (University of Washington)
