Rainfall from tropical cyclones: high-resolution simulations and seasonal forecasts

Wei Zhang, Gabriele Villarini, Gabriel Andres Vecchi, Hiroyuki Murakami

Research output: Contribution to journalArticlepeer-review

27 Scopus citations


This study examines the performance of the Geophysical Fluid Dynamics Laboratory Forecast-Oriented Low Ocean Resolution version of CM2.5 (FLOR; ~ 50-km mesh) and high-resolution FLOR (HiFLOR; ~ 25-km mesh) in reproducing the climatology and interannual variability in rainfall associated with tropical cyclones (TCs) in both sea surface temperature (SST)-nudging and seasonal-forecast experiments. Overall, HiFLOR outperforms FLOR in capturing the observed climatology of TC rainfall, particularly in East Asia, North America and Australia. In general, both FLOR and HiFLOR underestimate the observed TC rainfall in the coastal regions along the Bay of Bengal, connected to their failure to accurately simulate the bimodal structure of the TC genesis seasonality. A crucial factor in capturing the climatology of TC rainfall by the models is the simulation of the climatology of spatial TC density. Overall, while HiFLOR leads to a better characterization of the areas affected by TC rainfall, the SST-nudging and seasonal-forecast experiments with both models show limited skill in reproducing the year-to-year variation in TC rainfall. Ensemble-based estimates from these models indicate low potential skill for year-to-year variations in TC rainfall, yet the models show lower skill than this. Therefore, the low skill for interannual TC rainfall in these models reflects both a fundamental limit on predictability/reproducibility of seasonal TC rainfall as well as shortcomings in the models.

Original languageEnglish (US)
Pages (from-to)5269-5289
Number of pages21
JournalClimate Dynamics
Issue number9-10
StatePublished - May 1 2019

All Science Journal Classification (ASJC) codes

  • Atmospheric Science


Dive into the research topics of 'Rainfall from tropical cyclones: high-resolution simulations and seasonal forecasts'. Together they form a unique fingerprint.

Cite this