Twenty-first-century projections of North Atlantic tropical storms from CMIP5 models

Assessing potential changes in North Atlantic (NA) tropical storm (TS) activity this century is of paramount societal and economic significance, and the topic of intense scientific research(1). We explore projections of NA TS changes over the twenty-first century by applying a statistical downscaling methodology(2,3) to a suite of experiments with the latest state-of-the-art global coupled climate models(4). We also apply a methodology(5) to partition the dominant sources of uncertainty in the TS projections. We find that over the first half of the twenty-first century radiative forcing changes act to increase NA TS frequency; this increase arises from radiative forcings other than increasing CO2 (probably aerosols). However, NA IS trends over the entire twenty-first century are of ambiguous sign. We find that for NA TS frequency, in contrast to sea surface temperature (SST), the largest uncertainties are driven by the chaotic nature of the climate system and by the climate response to radiative forcing. These results highlight the need to better understand the processes controlling patterns of SST change in response to radiative forcing and internal climate variability to constrain, estimates of future NA IS activity. Coordinated experiments isolating forcing agents in projections should improve our understanding, and would enable better assessment of future TS activity.