The figure shows an average number of tropical storms, tropical cyclones and intense tropical cyclones per year simulated by the 0.23^ox0.31^o version of CAM5.1 for the four US CLIVAR HWG idealized AGCM configurations. Both increased atmospheric CO2 and elevated sea surface temperature reduce the total number of tropical cyclones. Error bars represent 5%-95% confidence intervals based on interannual variability.

We have found that the 0.25^o version of the CAM5.1 model reproduces global tropical cyclone statistics very well. Using the TECA tropical cyclone tracking tool, we have analyzed both AMIP-style simulations of the period 1979-2005 as well as the idealized numerical experiments of the US CliVar Hurricane Working Group. Highlights of this line of research include:

• Validation that the model produces the correct number of annual tropical cyclones globally.
• Validation that the model produces intense hurricanes (Category 4 and 5).
• Validation that the relationship between minimum pressure and maximum wind for simulated tropical cyclones follows that from observations.
• Projection that a warmer world will experience fewer total tropical cyclones but that the Cat 4 and 5 storms of will increase in frequency and severity.
• Concluded that inferences about future tropical cyclone activity from CMIP5 class models must be treated with extreme caution due to limitations from their coarse horizontal resolution.
• Shared data with the US CliVAR HWG resulting in 7 published papers. Others are to follow.

Source: Wehner et al. (2014) Resolution dependence of future tropical cyclone projections of CAM5.1 in the US CLIVAR Hurricane Working Group idealized configurations. To appear in J. Climate.