Testing the performance of tropical cyclone genesis indices in future climates

7 Tropical cyclone genesis indices (TCGIs) are functions of the large-scale environment which are 8 designed to be proxies for the probability of tropical cyclone (TC) genesis. While the perfor9 mance of TCGIs in the current climate can be assessed by comparison to observations of TC 10 formation, their ability to represent future TC activity based on projections of the large-scale en11 vironment cannot. Here we examine the performance of TCGIs in high-resolution climate model 12 simulations of current and projected climates, with a particular interest in determining whether 13 the index, when derived from the climatological seasonal cycle and spatial distribution of both TC 14 genesis frequency and large-scale fields from present climate, but then computed from large-scale 15 fields taken from simulations forced with SST patterns derived from coupled simulations of future, 16 warmer, climate scenarios can capture the global mean decreases in TC frequency found in those 17 future scenarios. This decrease is captured only when the humidity predictor is column saturation 18 deficit (the difference between actual and saturation water vapor) rather than relative humidity (the 19 ratio of these quantities). Using saturation deficit with relative SST as the other thermodynamic 20 predictor over-predicts the TC frequency decrease, but using potential intensity as the thermody21 namic predictor gives a good prediction of the decrease’s magnitude. These positive results appear 22 to depend on the spatial and seasonal patterns in the imposed SST changes; none of the indices 23 captures correctly the frequency decrease in simulations in which the only climate forcings are 24 spatially uniform, whether a globally uniform increase in SST of 2K, or a doubling of CO2 with 25 no change in SST. 26