Seasonal Cycles in North Atlantic and Western Pacific Tropical Cyclone Maximum Intensity

Hurricane activity in past decade has illustrated the destructive capabilities of powerful tropical cyclones (TCs). A useful theoretical framework, known as potential intensity (PI), allows researchers to constrain climatological maximum TC intensities given environmental conditions. A recent study by the authors showed that PI seasonality averaged over the North Atlantic (NA) main development region has a large seasonal amplitude that is dominated by SST seasonal cycles. In contrast, the Western North Pacific (WNP) PI seasonal cycle is relatively flat, permitting powerful TCs any month of the year, in part because near-tropopause TC outflow temperatures damp WNP PI seasonality by 30-40%. But how relevant are these climatological PI seasonal cycles for real-world TCs? This study compares the seasonal cycles of observed TC maximum intensities to those of PI. To directly compare, our study accounts for seasonal differences in TC track and frequency. TC best-track data is used to develop an along-track record of climatological PI for each observed storm in the satellite-era (1980-2015). When observed wind speeds are normalized by PI, the resultant distributions are approximately uniform; this property implies that PI theory is expected to hold in any month and basin with at least 25 observed historical storms. Results show that seasonal cycles of along-track PI and observed lifetime maximum intensity are similar to main development region averages: WNP TCs have observed maximum intensities near PI, are seasonally damped, and are observed in nearly every month of the year. In contrast, intense NA TCs are confined to the classical hurricane season and have a larger amplitude intensity seasonal cycle. A log-additive model shows that along-track SSTs remain dominant in seasonal variations in NA maximum intensity, whereas TC outflow temperatures are the leading contributor to WNP seasonal cycles of along-track PI. Finally, we use a combination of Monte-Carlo and Kolmogorov-Smirnov statistical methods to further investigate the uniformity of normalized TC intensities found in each TC basin.