I reviewed and compared the results of the NOAA AOML Maximum Sustained Wind Swath model against actual windspeed and gust information reported in the NHC Katrina advisories from 8am and noon on August 29th, as Katrina made landfall along the northern Gulf Coast.

This sparse dataset consists of readings from five different locations, plus the NHC estimate of max sust winds at noon, compared with the model estimate (to get an inland number, since I didn't find anything for Hattiesburg). The readings were from five different NWS anemometers, so we can assume they were fairly accurate.

In all six of these, the max sustained wind identified by the model was significantly lower than the actual reading, by an average of 23-27mph. Roughly, the error worsened on the east side of the storm, and the closer to the eye.

The maximum windspeed on land (not barrier islands) the model determined was 105mph, or a strong Cat 2, and that only for a very very small portion of the coastline, approximately from Gulfport to Biloxi.

However, note that the sustained windspeed at Gulfport was already 94mph, sometime between 7am and 8am, and this was still three hours from the closest approach of the eyewall to that city. The center of the eye was around 75 miles away from Gulfport at the time of that observation, when at its closest it would be less than 35 miles away, around 10:30am.

The actual data is attached.

Here are the referenced URLs:

http://www.aoml.noaa.gov/hrd/Storm_pages/katrina2005/wind.html

http://www.nhc.noaa.gov/archive/2005/KATRINA.shtml?

Not only are actual measurements quite different from the model output, but I have the eyewitness account of my brother which is also quite different.

My brother called me at 11am on the 29th and said he'd had sustained winds over 100mph for four hours; that is, starting around 7am. Two actual readings that support this.

First, he was 4 miles NNE of the eastern tip of Biloxi, close to the area of highest winds, which was Gulfport/Biloxi area. But he was 2 miles from water (Back Bay of Biloxi). Sometime between 7am and 8am, Pascagoula, which was to his ESE, almost 20 miles further away from the eye, reported to NHC a max wind gust of 118mph. Second, at noon, Dauphin Island, which was an additional 25 miles ESE, for a total of 45 miles to his east, a considerable distance wrt a hurricane windfield, reported sustained winds of 76mph, which would have undoubtably been higher an hour earlier, at 11am.

My brother said that around noon they had 90 minutes of extremely strong sustained winds and much darker lower clouds, and that during this time the wind direction swung around on the compass, so he knew he was close to the eyewall. He is not 100% certain about exactly when this 90 minutes occured. Hurricane hunters at the Biloxi Colliseum 5 miles to his west reported seeing a brief glimpse of lighter cloud cover as the eye passed to the west of them and this was previously on the web, but not anymore, and I believe was also around noon.

It is extremely likely he was in the eyewall winds. The model graphic generally confirms this. However the model generally seems to diminish the winds too rapidly at landfall, and also seems to overdo the reduction of winds due to friction with land.