Here is Judith Curry’s review.
Our understanding of the climate variability of tropical cyclone intensity is
hampered by lack of an insufficient historical data record and the inability of
current climate models to resolve tropical cyclones. Hence efforts are being
made to develop proxies for hurricane activity for which we have adequate
historical data or that can be resolved by climate models, to help us understand
how and why hurricane activity has varied in the past and how it might change in
As a proxy for changes in hurricane intensity, the Vecchi and Soden paper
examines the tropical cyclone maximum potential intensity. Maximum potential
intensity is a theoretical upper limit on hurricane intensity based on sea
surface temperature and the local vertical thermodynamic structure of the
atmosphere. There have been several theories of maximum potential intensity
proposed, and the most widely accepted of these is by Kerry Emanuel. Because
historical data on the vertical thermodynamic structure of the atmosphere are
not available prior to 1950, Vecchi and Soden further develop a proxy for
potential intensity based upon the tropical sea surface temperature anomalies
relative to the average tropical sea surface temperatures.
The main result of Vecchi and Sodens analysis is that maximum potential
intensity does not correlate well with the global increase in tropical sea
surface temperature, but rather with the regional variations in the surface
temperature. Looking specifically at the North Atlantic, the historical data
record since 1880 shows an increase in the Atlantic potential intensity that is
strongest near the equatorial latitudes in the eastern Atlantic, while the
Caribbean and Gulf of Mexico show a decrease in potential intensity. In terms
of the climate model projections for the North Atlantic, the maximum potential
temperature increases towards the equator and the African coast and also in the
Gulf of Mexico, but decreases in a swath extending from the Caribbean northeast
to the African coast north of about 25o.
So what is the significance of this paper in assessing the impact of global
warming on hurricane intensity, and particularly for the North Atlantic? North
Atlantic hurricane intensity has shown a strong relationship to sea surface
temperature, and the theory of maximum potential intensity has been used as a
physical link between increasing tropical sea surface temperatures and
increasing hurricane intensity. This paper is one of a number of papers
published in the last few years indicating that the influence of increasing sea
surface temperature on hurricane activity is complex; it is not just the local
heating effect of warmer sea surface temperatures (the maximum potential
intensity argument), but changing gradients of sea surface temperature influence
atmospheric circulations that can influence hurricanes. The eastern Atlantic
has been warming more rapidly than the western Atlantic (Gulf of Mexico and
Caribbean) which is influencing the atmospheric circulations and resulting in
more hurricane genesis in the eastern Atlantic and more equatorward tracks.
This paper is an interesting contribution that will be of some help in filling
out the picture of how climate variations influence hurricanes. There are some
caveats about the methods used in the paper, that will hopefully be addressed in
future research; these caveats are:
* Current theories of hurricane maximum potential intensity still fall short in
terms of accurately accounting for the exchange of heat between the atmosphere
and the ocean and the asymmetric vortex dynamics.
* The SST data is not of adequate quality prior to 1920 especially in the
Pacific Ocean, particularly for the proxy regional SST anomalies.
* Climate model projections of potential intensity depend on the
parameterization of tropical convection to provide the upper atmospheric
temperature and humidity used in the calculation of potential intensity.
Tropical convection parameterization is one of the weakest links in climate
models. Confidence in the climate model projections of potential intensity
should have been established by looking at the climate model simulations for the
past century and comparing with the historical data record of potential
* Climate models generally do a poor job of simulating realistic interannual
and decadal modes of natural climate variability. The climate models used here
show strong interannual and decadal variability in maximum potential intensity;
assessing this variability using climate model simulations of the last 100 years
would have made the authors claim about the magnitude of the natural
variability exceeding the trend more credible.