Network News

X My Profile
View More Activity
The new Washington
Post Weather website
Jump to CWG's
Latest Full Forecast
Outside now? Radar, temps
and more: Weather Wall
Follow us on Twitter (@capitalweather) and become a fan on Facebook
Posted at 9:30 AM ET, 09/ 2/2008

Computer Models Nailed Gustav's Track

By Jason Samenow

A Gustav forecast from the GFDL model simulated 5.5 days prior to landfall (left). A satellite image of the actual Gustav (right). Notice how well the model approximated Gustav's location.

Gustav Resources: WaPo: Levees Tested | Storm Tracking | Photo Gallery | New Orleans Times-Picayune

Kudos to the Geophysical Fluid Dynamic Laboratory (GFDL) weather model. It provided an amazingly accurate forecast of Hurricane Gustav's track more than FIVE days before landfall (recall I showed this last Thursday, and compared it to Katrina). The above image shows the similarity between Gustav's five-day model simulation and a satellite image of the real Gustav just prior to coming ashore. The model was about 12 hours too fast, but gradually got the timing right in subsequent simulations.

Congratulations are also in order to the National Hurricane Center who pegged southeast Louisiana in the middle of its track guidance 5-6 days before landfall. Watch this remarkable animation of the Hurricane Center's evolving track forecast from the time it began issuing advisories on Gustav. Notice how consistently accurate the forecast was for the landfall location.

Keep reading for an assessment of how well Gustav's intensity forecast was handled. For Washington, D.C. weather, see our full forecast and NatCast for tonight's game.

The intensity forecast by some of the models (like the GFDL), while by no means terrible, wasn't as stellar as the track forecast. You'll notice in the GFDL simulation above, Gustav is much more symmetric than the real Gustav. The model forecast a pressure of about 935 mb and maximum winds of 140 mph (Category 4) compared to 954 mb and maximum winds of 115 mph for the real thing. Some of the models did not capture the effect of the wind shear which disrupted Gustav's structure and prevented it from strengthening over the warm Gulf of Mexico. Chief meteorologist Jeff Masters at wunderground provides an excellent technical explanation of how the structure was impacted:

...passage over Cuba disrupted the eyewall structure just enough to allow the upper-level winds shearing it to penetrate into the heart of the hurricane. These winds ripped up the eyewall and tilted it, so that the surface eye was no longer underneath the upper-atmosphere eye. A tilted eyewall structure is not able to act as an efficient heat engine until it can get itself lined up more vertically, so Gustav was unable to take advantage of the warm Loop Current waters it was traversing. It's like when your car engine is not firing on all cylinders and you hit the gas pedal--nothing happens. Once Gustav finally did align its eyewall vertically and armored itself against the effects of the wind shear, it had passed beyond the Loop Current and was over cooler waters of much lower heat content. Thus, Gustav was not able to intensify much before landfall. The computer models that predicted a Category 4 hurricane at landfall could easily have been correct, had the shear been a few knots less when Gustav crossed Cuba.

The forecasting experience with Gustav reinforces what we already know: track forecasts are pretty good (but by no means perfect) while intensity forecasts are much more difficult. I think we'll see that again with Hurricane Hanna...which we'll be updating you on in just a couple hours...

By Jason Samenow  | September 2, 2008; 9:30 AM ET
Categories:  Tropical Weather  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: Forecast: Some Summer in September
Next: Could Hanna Hit D.C. Hard?


NHC has Hanna tracking very close to us between 8 AM and 1 PM Saturday, with a 10%-20% chance of tropical storm-force winds. [While this is not a "big" dance weekend for me, Arlington County has a jazz festival scheduled from noon to 7 PM Saturday in Rosslyn. This event could be DEFINITELY IMPACTED if Hanna decides to act up locally! By 8 PM Hanna will be up in New England, so the storm will be a fast mover. However Tucker Barnes @ Fox 5 is talking "good soaking rain" for the day Saturday.]

Josephine just formed somewhere out near Cape Verde about half an hour ago. Ike is in between Hanna and Josephine in the central Atlantic.

Posted by: El Bombo | September 2, 2008 11:44 AM | Report abuse

Jason, not only did you show this forecast track last week, you were excoriated for the headline above it. I would have hoped that those who were quick to criticize you would be just as quick to come back and admit that not only was the GFDL right, but you were also right in choosing that model as the basis for a Gustav post.

Posted by: ~sg | September 2, 2008 1:08 PM | Report abuse

Great post; thanks Jason!

Posted by: Artjohn | September 2, 2008 1:28 PM | Report abuse

Gustav's eventual track was extremely well modeled compared to many storms. For several days the majority of models took it to central LA.

Posted by: Ian, Capital Weather Gang | September 2, 2008 5:05 PM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2012 The Washington Post Company