Estimating the Causes of Past Atlantic Tropical Cyclone Multidecadal Variability

Abstract:

Atlantic Tropical Cyclone (TC) activity exhibits large variability over a wide range of timescales, including multidecadal timescales. Given the limitations of climate models in simulating future TC activity, understanding the causes of past variability is important, especially on long timescales. To explain causes of past multidecadal variability, we focus on the marked TC activity decrease of the 1970s and 1980s. TC activity depends strongly on the thermodynamic environment in which TCs develop, and potential intensity (PI) is a variable that expresses well how favorable an environment is to TC activity. Hence, here we focus on explaining the past variability of PI itself. Using a novel framework based on the weak temperature gradient approximation and single column model simulations, we partition historical changes to the basin PI between contributions from local and global forcings. This framework allows to estimate and compare the effects of the variability of local forcings such as dust aerosols, surface winds, or ocean heat flux convergence, on the variability of PI. Within that framework, we find that local forcings dominate PI variability, and that half of the historical mutidecadal PI variability can be explained by variations in dust emissions from the Sahara-Sahel. It has been shown that, in the Sahara-Sahel, dust emissions correlate well with drought conditions, which were amplified in the 1970s and 1980s by European and American SO2 emissions. Thus, past TC variability has likely been influenced indirectly by anthropogenic forcing, via its effects on the climate-sensitive Sahel.