A Comparison of Tropical Cyclone Projection in a High-Resolution Global Climate Model and Downscaled by Statistical and Statistical-deterministic Methods

Abstract:

In this study, we investigate the response of tropical cyclones (TCs) to climate change by using the statistical-deterministic downscaling method of Emanuel et al. (2008) and the Princeton environment-dependent probabilistic tropical cyclone model (PepC) of Jing and Lin (2020) to generate large numbers of TCs, with environmental conditions taken from the Geophysical Fluid Dynamics Laboratory (GFDL) High-Resolution Forecast-Oriented Low Ocean Resolution (HiFLOR) model for both the historical period and a future scenario under the Representative Concentration Pathway 4.5. The downscaled TCs under the current climate are compared with those in the mid-21^st and late-21^st century for each downscaling technique as well as with TCs explicitly simulated in HiFLOR. We show that while more storms are detected in HiFLOR towards the end of 21^st century, TCs downscaled with the statistical-deterministic model shows a moderate increase in TC total frequency, and TCs downscaled with PepC shows no significant change. The changes in storm frequency for all three datasets are not significant in the mid-21^st century. In all three datasets, storms become more intense and there is significant increase in the frequency of major hurricanes and Category 5 storms in the future climates. However, HiFLOR has the largest increase in intensity while TCs downscaled with PepC has the least. Storm intensification leads to small changes in the landfall intensity distribution in all three datasets. However, larger changes and variations are found in regional storm intensity return levels. TC intensity return levels in the North-East USA grow the fastest as the climate gets warmer, though the region has the smallest TC threat under the current climate. On the contrary, in Mexico the TC threat may even slightly decrease under a warmer climate.