High-resolution (i.e., grid spacing less than 50 km) climate models are now becoming promising tools to evaluate hurricane characteristics in past, current, and future climate conditions. The Community Atmosphere Model version 5 (CAM5) at high horizontal resolutions (~28 km grid spacing) has been used to study the impact of future climate change on global tropical cyclone (TC) activity. Previous CAM5 analysis has traditionally focused on evaluating climate controls on TC frequency, duration, and intensity.

The goal of this work is to better understand climate impacts on TC precipitation and size in various configurations of high-resolution CAM5 run at horizontal grid spacings of approximately 28 km forced with prescribed sea-surface temperatures (SSTs) and greenhouse gas concentrations for past, present, and future climates. This analysis will include the evaluation of TC characteristics in conventional (AMIP-style) decadal simulations typical of climate models, as well as in short 7-day ensemble hindcasts of recent devastating events (e.g., Hurricane Irma in 2017). These hindcast simulations are initialized with atmospheric analyses from NOAA’s Global Forecast System (GFS) before the time of interest (e.g., landfall) for each individual storm, and the ensemble is created through parameter perturbations in the CAM5 physics parameterizations to provide a sufficient sampling of uncertainties in storm characteristics. This ensemble of hindcasts is referred to as “the world that was” simulations. A second suite of ensembles are run using alternate initial conditions where the large-scale climate change signal has been removed, a “world that could have been” scenario. Through comparison of the “world that was” and “world that could have been” ensembles for each individual storm, the impact of climate change on individual TC rainfall and size will be quantified and compared to the full decadal-style simulations that analyze TCs using interannual statistics over hundreds of TCs.