At 5 PM EDT, the center of Hurricane Katrina was at 24.8 North 82.9 West, 70 miles west-northwest of Key West. Maximum sustained winds remain near 100 mph and minimum central pressure is down to 965 millibars (28.50").
Hurricane Katrina Advisory Number 14
Despite the recent drops in pressure, recon has not found flight level winds that would suggest anything greater than 100 mph winds at the surface. The fact that the eyewall is not entirely closed may be the reason why winds are a bit lower than the pressure would normally suggest.
Katrina remains caught between steering flows that are pushing her south. Both sources of that are expected to weaken over the next 24 hours, causing Katrina to turn to the west. The models are in general agreement on the movement of a trough that will come from the northwest of the high pressure ridge that will act to open up a path for Katrina to turn north. The result of the better agreement of the models on the scenario caused the official track forecast to be shifted to the west by 150 nautical miles (172.5 statute miles). With projected landfall still 72 hours away, further modifications to the forecast track are possible.
Katrina is expected to move over an area of exceptionally warm (and deep) water. That combined with favorable upper-air conditions should allow Katrina to reach category four status before landfall. The intensity forecast is in line with the output of the SHIPS and GFDL models, but under that of the FSU Superensemble, which brings Katrina's winds up to 150 mph strength prior to landfall.
Hurricane Katrina Discussion Number 14
Official forecast track
A very difficult move to make, but probably the right one. The National Hurricane Center really does not like to make these kinds of jumps in their forecast, but the unaminous shift in the forecast models combined with Katrina's continued movement to the west-southwest left them little choice.
The only direct comment I have on the forecast has to do with intensity. After looking at the water vapor loop, I realized dry air could serve to keep Irene from doing dramatic rapid intensification. Right now there is dry air immediately to her north and there looks to be a pocket of it being pushed south through the southeastern U.S. and will ultimately end up in Katrina's path. (In these imagery, dry air is black, while increasingly bright shades of gray indicate moisture). It may not amount to much, Katrina probably will still manage to make category four status. However, if she doesn't then dry air will probably be the reason why.
Brendan Loy has a post about the peril that New Orleans may be in and the difficult problems of the track forecast and evacuation orders. It is indeed a difficult problem.
The forecasting of the tracks of hurricanes is much more accurate than in the past and overall improves every year. However, we are still far from being able to give the kind of precision and certainty that would alleviate the emergency planner's fear of a false alarm.
The problem in New Orleans is that it is generally accepted that the city requires 72 hours notice to be properly evacuated. In 2004, a great year for the accuracy of hurricane forecasts relative to the historical average, the average track error of a 72 hour forecast was 150 nautical miles. Given that error, a forecast that has a hurricane striking east of Mobile in 72 hours(well enough away to not be of worry to New Orleans if it were indeed to strike there), still has enough error that New Orleans is inside the zone of uncertainty.
Even at a short period the uncertainty can be greater than would be desired by planners. In 2004 the average forecast error at 12 hours was 33 nautical miles. Now this sounds small but in practical terms it is somewhat large. Consider that in 1969, when the monsterous Hurricane Camille passed New Orleans to the east by about 40 nautical miles the damage to New Orleans was trivial compared to what it would have been had it come in 30 miles closer. The difference between absolute disaster, major damage (like Betsy of '65) or annoyance is within the margin of forecast error at 12 hours.
As much as forecasters would like to, they can't give the assurances that emergency planners would like.