Gustav pulls away from Hispaniola after killing at least 22

Tropical Storm Gustav continues to pound Haiti with torrential rains and tropical storm-force winds, as the storm slowly moves westward towards Jamaica and Cuba. Flooding form Gustav has killed at least 22 people--14 on Haiti, and 8 in the Dominican Republic. Satellite rainfall estimates show that many portions of the southern peninsula of Haiti have received 8 or more inches of rain, and an additional 2-4 inches are likely over the worst-hit regions before Gustav finally moves away. The encounter with Haiti's high mountains has weakened Gustav. The latest center fix from the Hurricane Hunters at 3:34 pm EDT put Gustav's central pressure at 999 mb, and the aircraft was not able to find surface winds above 55 mph.


Figure 1. Current satellite image of Gustav.

Visible satellite loops and radar from Gran Piedra, Cuba show that Gustav starting to become better organized, with an increase in heavy thunderstorm activity and better spiral banding. The pressure should start to fall tonight, and the winds should increase to hurricane strength by Thursday night.

The track forecast for Gustav
The latest 12Z (8 am EDT) model runs show not much change to the short term track of Gustav, but there is a major long-term forecast change to report. In the short term, the models agree on a track through the narrow channel between Cuba and Jamaica. However, the HWRF model takes Gustav over Jamaica, and residents there should anticipate the possibility of a Category 1 hurricane strike on Thursday. However, heavy rain is probably the primary threat to Jamaica and eastern Cuba. The storm is small enough that high winds will be confined to a small region around the center, and a near miss will not cause much wind damage.

A ridge of high pressure is expected to force Gustav west through Friday. By Saturday, a trough of low pressure moving across the Midwest U.S. should weaken the ridge, and allow Gustav to turn north across western Cuba into the Gulf of Mexico. The final landfall location of Gustav depends on the strength and speed of the trough. In this morning's runs, we now have some indication that this trough of low pressure may not be strong enough to pull Gustav to the coast by Monday. In fact, the NOGAPS model predicts Gustav will stall offshore the Alabama coast on Monday, before finally edging ashore three days later. The UKMET model is also much slower with its latest run, and slows Gustav down as it approaches the Louisiana/Texas coast on Monday. The latest HWRF model is also slower than the last run, and doesn't bring Gustav to the coast by the end of its forecast period (Monday). The HWRF foresees a Category 3 or 4 hurricane a few hundred miles south of the Louisiana coast on Monday. The latest ECMWF run is not much slower than the previous run, but does stall Gustav out over central Louisiana once it makes landfall near New Orleans Monday night. The latest GFDL model, though, is not much slower, and predicts landfall in Mississippi early Monday morning as a Category 3 hurricane. In summary, Gustav may slow down considerably just before landfall in the U.S., making its long-term track and landfall location very uncertain at this time.

The intensity forecast for Gustav
Now that Gustav is clearing the southwest Peninsula of Haiti, it should begin to strengthen. Gustav is currently under low wind shear (5-10 knots). This shear is expected to remain in the low to moderate range (0-15 knots) for the remainder of the week. By Friday, as Gustav approaches the Cayman Islands, the storm will be underneath an upper level anticyclone, and over the highest heat content waters of the Atlantic. The GFDL model responds by intensifying Gustav to a Category 3 hurricane as it passes through the Caymans. I think it is plausible that Gustav could intensify further, to Category 4 strength, before hitting the Caymans, though the most likely intensity is Category 2. Keep in mind that our skill forecasting intensity changes is poor. If you have plans to be on the northern Cayman Islands--Cayman Brac or Little Cayman Island--on Friday, be prepared to be stuck there for several days, as Gustav may heavily damage these islands. Grand Cayman Island is also at risk--the HWRF model predicts Gustav will pass over Grand Cayman on Friday afternoon.

Gustav will lose intensity when it crosses the western tip of Cuba--perhaps by 25 mph or so--but should easily regain that lost strength within 12-24 hours. Gustav will likely be a major Category 3 or 4 hurricane in the Gulf of Mexico. However, Gustav not be able to maintain that strength all the way to landfall, if it slows down close to landfall, as some of the models are predicting.

Gustav's impact on Cuba and Mexico's Cancun/Cozumel region
If you're in the cone of uncertainty, you're at risk. Hurricane forecasts are uncertain, and this uncertainty is graphically represented by the cone of uncertainty around the central "best guess" forecast positions. Cancun and Cozumel are both at the edge of the cone of uncertainty, so could get a direct hit from Gustav. At present, though, it appears that Mexico's Yucatan will only have one day of heavy rain (Saturday) with some winds gusts of 40 mph. The odds of Cozumel getting sustained winds of tropical storm force (39 mph or greater) are 20%, as indicated in NHC's wind probability product. The odds of hurricane force winds are 9%.

The portion of the world most likely to suffer a major hurricane strike from Gustav will be western Cuba. Gustav is likely to make landfall as a major hurricane somewhere along this stretch of coast, bringing a storm surge of 10-15 feet to the right of where the eye come ashore. Gustav may bring hurricane-force winds to the capital, Havana (on the north shore), which would cause heavy damage to the many poorly-built structures in the city.

Gustav's potential impact on the oil and gas industry
The price of U.S. crude oil has jumped about 2%, and the price of U.S. natural gas has increased 11% in the past two days, in anticipation that Gustav might rip through the oil and gas production areas of the Gulf of Mexico. About 25% of U.S. crude oil and 15% of its natural gas are produced in the Gulf of Mexico. As seen in Figure 3, the oil production areas are concentrated along the Louisiana and Texas coast. If Gustav makes a landfall on the right side of its cone of uncertainty, in Alabama or the Florida Panhandle, the oil and gas infrastructure might not be significantly affected. However, most of the cone of uncertainty lies in the major oil and gas producing areas, and I give a 60% chance that Gustav will significantly hurt oil and gas production in the Gulf.


Figure 2. Location of major oil platforms in the Gulf of Mexico. The data was taken from the Department of Interior Mineral Management Services

Gustav's intensification potential in the Gulf of Mexico
As we saw in 2005 with Katrina and Rita, the large amounts of deep, warm water brought into the Gulf of Mexico by the Loop Current can help intensify hurricanes to Category 5 intensity. As explained in my Loop Current tutorial, the Loop Current is an ocean current that transports warm Caribbean water through the Yucatan Channel between Cuba and Mexico. The current flows northward into the Gulf of Mexico, then loops southeastward through the Florida Keys. The Loop Current commonly bulges out in the northern Gulf of Mexico and sometimes will shed a clockwise rotating ring of warm water that separates from the main current. This ring of warm water slowly drifts west-southwestward towards Texas or Mexico at about 3-5 km per day. This feature is called a "Loop Current Ring", "Loop Current Eddy", or "Warm Core Ring", and can provide a key source of energy to fuel rapid intensification of hurricanes that cross the Gulf. The Loop Current itself can also fuel rapid intensification, such as happened with Hurricane Charley in 2004. When a Loop Current Eddy breaks off in the Gulf of Mexico at the height of hurricane season, it can lead to a dangerous situation where a vast reservoir of energy is available to any hurricane that might cross over. This occurred in 2005, when a Loop Current Eddy separated in July, just before Hurricane Katrina passed over and "bombed" into a Category 5 hurricane. The eddy remained in the Gulf and slowly drifted westward during September. Hurricane Rita passed over the same Loop Current Eddy three weeks after Katrina, and also explosively deepened to a Category 5 storm.

This year, we had another Loop Current Eddy break off in July. This eddy is now positioned due south of New Orleans (Figure 2), and this eddy has similar levels of heat energy to the 2005 eddy that powered Katrina and Rita. Should Gustav pass over or just to the left of this eddy, we can expect the storm to significantly intensify. There is also a weaker eddy present in the western Gulf; this eddy broke off from the Loop Current in April, and is much cooler then the eddy that broke off in July. Should Gustav pass over the April eddy, it shouldn't make much difference.

A NOAA Hurricane Hunter research aircraft is scheduled to drop a network of 20 specialized buoys (Air eXpendable BathyThermographs, or AXBTs) in the Gulf of Mexico Thursday to provide precise measurements of ocean temperatures in order to aid intensification forecasts for Gustav. This data will feed directly into the GFDL and HWRF computer models, but not into the other global models (GFS, NOGAPS, UKMET, and ECMWF).


Figure 2.Tropical Cyclone Heat Potential (TCHP) for the Gulf of Mexico on August 24, 2008. TCHP values in excess of 80 kJ/cm^2 (yellow colors) have been found to promote rapid intensification of hurricanes. The Loop Current is shown by the arrows at lower right, and begins in the Caribbean, flows north through the Yucatan Channel into the Gulf of Mexico, then loops back to the south and turns eastward over the Florida Keys. Two Loop Current Eddies have broken off from the Loop Current this year--one in April, and one in July. These eddies have drifted slowly westward, and still maintain heat from the Loop Current. Image credit: NOAA/AOML.

Links to follow
Wundermap for Cuba/Haiti/Jamaica
Gran Piedra, Cuba radar

Disturbance 95L east-northeast of Puerto Rico
A tropical disturbance (95L) near 19N 57W, a few hundred miles east-northeast of Puerto Rico, has become better organized today. Visible satellite loops show an increase in heavy thunderstorm activity, and the center of circulation has jumped about 100 miles to the southeast to be underneath the heaviest thunderstorms. This new location is also in a region of lower wind shear, and also means the latest 12Z (8 am EDT) computer model runs are likely to not be so good. Wind shear is a moderate 10-20 knots over 95L today, but is forecast to decrease below 5 knots on Thursday and remain below 15 knots for most of the remainder of the week. Dry air should not be a problem for 95L. NHC is giving 95L a medium (20%-50% chance) of developing into a tropical depression by Friday. I believe it is very likely (>70% chance) that this will be a tropical storm sometime in the next 2-5 days. Most of the models develop 95L, foreseeing a northwest track and a threat to Bermuda 4-6 days from now. The latest (8 am EDT) GFDL model run develops 95L into a strong Category 1 hurricane that passes very close to Bermuda on Sunday. The HWRF model predicts a Category 1 hurricane near Bermuda on Monday. In the longer range, it is possible that 95L will stall in the Bermuda area and move slowly, as steering currents collapse early next week. The storm may then get forced westward over Florida late next week, as forecast by the ECMWF model. However, the GFS and NOGAPS models are taking 95L northwards towards Canada late next week. It is too early to guess which of these solutions is more likely. There are currently no Hurricane Hunter missions scheduled into 95L.

Elsewhere in the tropics
Most of the computer models forecast the development of at least one additional tropical wave between Africa and the Lesser Antilles Islands during the coming week. The first candidate is a large circulation located near 17N, 34W, a few hundred miles west of the Cape Verde Islands. Wind shear is a moderate 10-15 knots in the region, and forecast to remain in the moderate to marginal range (10-20 knots) for the next few days. NHC is giving this system a low (<20% chance) of developing into a tropical depression by Friday, and some slow development is possible over the next few days. Only the GFS model develops it, and predicts the system will be near the northern Lesser Antilles Islands Monday or Tuesday.

My next blog will be Thursday morning.

P.S., a little humor from The Onion might be what is needed for some. Check out this morning's video, Hurricane Bound For Texas Slowed By Large Land Mass To The South.

Jeff Masters