Atlantic hurricanes and climate change: Projection of a peak month in a future record hurricane season

This research addresses the question of how climate change could impact seasonal Atlantic tropical cyclone (TC) activity by the end of the 21st century. Specifically, how might a particularly active season, such as that of 2005, change in the future? Numerical simulations of TCs provide an ideal tool with which to test the sensitivity of TC intensity to modifications in the surrounding environment. However long-term simulations using global climate models (GCMs) require coarse resolution. In this study, GCM-predicted temperature and moisture anomalies are incorporated into a higher-resolution simulation performed with the Weather, Research and Forecasting (WRF) model. The goal of this study is to predict how an active Atlantic TC season, such as that of 2005, might be altered in a future climate affected by global warming.

Horizontal grid spacing of 18-km is used to simulate the entire Atlantic basin during the peak month of the 2005 season. A 4-member physics ensemble is run by varying the microphysical and boundary layer schemes. The number of named storms decreases by 28%, hurricanes by 33%, and major hurricanes by 21%, in the warming ensemble compared with the control ensemble. Despite the warming of the ocean surface, the thermodynamic efficiency of TCs is not significantly changed in the future atmosphere due to the fact that both the average inflow and outflow temperatures of the TCs are warmer. The magnitude of the decrease in future TC activity is somewhat sensitive to the physics parameterizations used and also to whether or not increased CO₂ is explicitly accounted for in the model radiation scheme. Precipitation associated with the simulated future TCs is larger than that in the control, both for spatially averaged and maximum point values.

On the 18-km grid, major storms do occur frequently with some central pressures consistent with category 4 hurricanes. However, grid spacing this coarse does not allow adequate representation of physical processes important to TC intensity. Previous studies, such as Chen et al. (2007) and Gentry and Lackmann (2009), have shown that grid spacing at or below 2 km is needed to simulate the full intensity of TCs, and that even below horizontal grid spacing on the order of 10 km, the sensitivity to spatial resolution is significant. Therefore, a subset of TCs identified on the 18-km grid are simulated at higher resolutions by spawning multiple vortex-tracking moving nests. By increasing the resolution even further, the sensitivity of these results to horizontal grid spacing can be assessed.