New paper finds climate models exaggerate predictions of tropical cyclones

A new paper published in The Journal of Climate admits that climate models are greatly exaggerating the likelihood that global warming increases tropical cyclone activity or intensity. According to the authors, "temperature trends in NCEP [the National Center for Environmental Protection database] are unlikely to be accurate, and likely drive spuriously positive TC [tropical cyclone] and PI [potential intensity] trends, and an inflated connection between absolute surface temperature warming and TC [tropical cyclone] activity increases." Indeed, as observations show no increase in global cyclone activity. 

Journal of Climate 2012 ; e-View

doi: http://dx.doi.org/10.1175/JCLI-D-12-00503.1

Impacts of Atmospheric Temperature Trends on Tropical Cyclone Activity

Gabriel A. Vecchi,¹ Stephan Fueglistaler,² Isaac M. Held,¹ Thomas R. Knutson,¹ and Ming Zhao^1,3

¹ NOAA/GFDL, Princeton, NJ 08540 USA

² Princeton University, Princeton, NJ 08540 USA

³ UCAR, Boulder, CO USA

Abstract

Impacts of tropical temperature changes in the upper troposphere (UT) and the tropical tropopause layer (TTL) on tropical cyclone (TC) activity are explored. UT and lower TTL cooling both lead to an overall increase in potential intensity (PI), while temperatures 70hPa and higher have negligible effect. Idealized experiments with a high-resolution global model show that lower temperatures in the UT are associated with increases in global and North Atlantic TC frequency, but modeled TC frequency changes are not significantly affected by TTL temperature changes nor do they scale directly with PI.

Future projections of hurricane activity have been made with models that simulate the recent upward Atlantic TC trends while assuming or simulating very different tropical temperature trends. Recent Atlantic TC trends have been simulated by: i) high-resolution global models with nearly moist-adiabatic warming profiles, and ii) regional TC downscaling systems that impose the very strong UT and TTL trends of the NCEP Reanalysis, an outlier among observational estimates. Impact of these differences in temperature trends on TC activity is comparable to observed TC changes, affecting assessments of the connection between hurricanes and climate. Therefore, understanding the character of and mechanisms behind changes in UT and TTL temperature is important to understanding past and projecting future TC activity changes. We conclude that the UT and TTL temperature trends in NCEP are unlikely to be accurate, and likely drive spuriously positive TC and PI trends, and an inflated connection between absolute surface temperature warming and TC activity increases.