A POSSIBLE CONNECTION BETWEEN GLOBAL WARMING

AND THE 2005 ATLANTIC HURRICANE SEASON?

Stu Ostro

The Weather Channel

December 2005

This is not meant to be passed off as a rigorous scientific study. It is just a sidebar to a blog entry … but is meant to provoke thought: if the general state of the atmosphere is changing, why should tropical cyclones be immune from being affected? Exactly how this might be happening is a controversial issue given the complexity of the system, issues with historical data, and the nature of individual events. The line of reasoning below – one which avoids citing SSTs or statistical trends – is presented as one way a physical connection might have existed during this exceptional season.

Note: All maps are from the NOAA/CDC reanalysis site and are for 24 August 2005, the date of Katrina’s development east of Florida, through 24 October, the date of Wilma’s U.S. landfall. In other words, the mid-season peak during which three of the lowest six pressures on record in the Atlantic basin occurred.

(The only exceptions are the shear maps, for which data was only available through 11 October, but they are consistent with those for the whole period. Also, graphics are not included, but results for the period going back through the time of Hurricane Dennis are similar.)

The period was characterized by anomalous warmth across much of the Northern Hemisphere (see below). Per NCDC, there and globally (land and ocean mean temp) October and September were the warmest on record going back to the late 1800s. August was the 3^rd warmest. July was the warmest in the Northern Hemisphere and 2^nd warmest globally.

In & near North America, there was a positive surface temperature anomaly centered from the north-central U.S. across eastern Canada, with a broad significantly positive anomaly in the entire eastern 2/3 or so of the United States.

During the period from mid-summer through mid-autumn (July through October), there were 954 individual daily record highs set in the U.S. (compared with 269 record lows).

There was a corresponding 500mb height anomaly:

And at 200mb:

In turn, there was a wind anomaly at 200mb associated with this exceptionally strong ridge, with lighter-than-average winds and less shear than average in key tropical cyclone formation areas.

200mb zonal wind anomaly:

200mb vector wind anomaly:

850-200mb wind shear anomaly:

This is reflected in a comparison of the mean shear during this period …

… with the long-term climatology, which shows pronounced W-NW 850-200mb shear over the subtropical Atlantic, Gulf, and Caribbean.

The mean 200mb flow for the same period in 2005, which shows a predominant upper-level anticyclone in the eastern Gulf, Caribbean, and subtropical Atlantic:

Which in turn corresponds to where many of 2005’s tropical cyclones went through their main development phase(s) – letters below denote approximate locations – including the most prominent ones (Dennis, Emily, Katrina, Ophelia, Rita, Wilma).

[A₂, B₂, , and E₂ denote Alpha, Beta, Gamma, Delta, and Epsilon]

So, does this prove a global warming connection i.e. without that there couldn’t have been a strong ridge oriented in a similar fashion and concurrent extremes in tropical cyclone intensity and number?

No; there are historical examples of very active times in terms of total number and severe U.S. impacts long before the warming signal had become evident and before the topic was the talk of the town – e.g. the 1890s. In the same vein, the maps did not look like those above for the same period in 2004, another active and destructive year (although 2004 did not have 28 storms and four Cat 5s…). The longwave pattern is driven by a myriad of other factors besides anthropogenic-induced forcing, and in general there’s more to the 2005 tropical cyclone story (other dynamics, instability, etc.) than what’s presented here.

That all having been said, however … if one accepts that anthropogenic warming contributed to the above-average temperatures across much of the hemisphere and in particular the anomalous heat/warmth in eastern North America, and that in turn was associated with the strength of the ridge aloft, then this would seem to at least possibly provide a means by which the changing climate – beyond just natural multidecadal variability – could have influenced the number, intensity, and/or tracks and genesis locations of the cyclones during this record season.