Caribbean wave develops a new circulation

A powerful tropical wave that has the potential to become a serious hurricane is sweeping through the southernmost Lesser Antilles Islands today, bringing tropical storm force wind gusts and heavy rains. An important development has occurred in the past two hours--a new circulation center developed near 12.5N 63W, about 100 miles north-northwest of the original center near the South American coast. This new center lies between St. Vincent and Grenada, and southwest winds observed last hour in Grenada confirms that a closed circulation now exists at the surface. The old center near the South American coast now looks likely to dissipate. Inflow of warm, moist air into its center was too restricted by the presence of the South American land mass, and thus a new center farther north along the axis of this tropical wave was able to form and take over.

Current conditions in the islands
Barbados reported sustained winds of 32 mph at 5am this morning, and wind gusts as high as 51 mph. Togabo had wind gusts to 36 mph, and sustained winds of 33 mph with wind gusts to 43 mph were observed on St. Lucia. A QuikSCAT satellite pass from 6:30am EDT shows a large area of 35 mph winds to the north of the center. Radar from Martinique shows the heavy bands of rain spreading over the islands. Winds on Martinique have gusted to 35 mph so far today.

Wind shear
What happens in the next few hours in crucial in determining if we have a serious hurricane to worry about in a few days, or just a another harmless tropical blob. The storm is very vulnerable to wind shear right now as it reorganizes. The center of circulation is almost completely exposed, with just one spiral band of heavy thunderstorms connected to the northeast side of the center. Upper level winds out of the west are creating about 10 knots of wind shear over the center, keeping the band of thunderstorms pushed to the downwind side of the center. If the shear can increase a little this afternoon, it may disrupt the storm enough to keep it from developing today. The new center location also puts the storm closer to the large area of dry air and Saharan dust that covers much of the eastern Caribbean. This may also help disrupt the storm.

However, I think 97L will overcome these obstacles. Wind shear is probably low enough to allow the storm to reform, and I wouldn't be surprised to see the center jump again more to the east, to be underneath the strongest thunderstorms. The shear is forecast to remain low through the next five days. There is a zone of very high shear to the system's north, but it is forecast to retreat to the west ahead of the developing storm.

The computer models
The latest 8pm and 2am EDT computer model runs were all initialized with the old center position, and thus are unreliable. The new set of model runs using data from 12 GMT (8am EDT) are not going to be any better, since the new center formed about 2 hours later. We'll have to wait until the 18 GMT (2pm EDT) model runs are available late tonight before we can put much stock in any of the computer model solutions. With this in mind, here is my summary from my previous blog about what the latest computer models say:

The Canadian model continues to be very consistent and very gung-ho, developing 97L into a strong tropical storm on Saturday, south of Jamaica, then taking the storm into the Gulf of Mexico as a hurricane. The NOGAPS model is also consistent, assuming a more southerly track will occur with no development due to close proximity to the South American coast. The GFS takes a weak tropical storm across the Dominican Republic on Saturday, then into the Bahamas. The GFDL has the same idea, but has a much stronger system that becomes a Category 1 hurricane in the Bahamas on Monday. The run-to-run consistency of the GFDL has been poor, and both the GFDL and GFS have not done a good job forecasting the initial track of the storm so far.

What the new center means for this storm
The separation of its center from the coast removes the primary impediment to intensification for 97L. It looks more likely that this storm will develop into at least a strong tropical storm, and probably a hurricane. The track such a hurricane might take is highly uncertain, but the more northerly center increases the risk for Haiti, the Dominican Republic, the Bahamas, and the U.S. East Coast.


Figure 1. Current satellite image of Invest 97L.


Figure 2. Preliminary model tracks for disturbance 97L approaching the Lesser Antilles Islands.

Debby
Tropical Storm Debby remains a minimal tropical storm today, but is expected to slowly intensify in response to some warmer waters along her path the next two days. By the time Debby turns north this weekend, she could attain hurricane status. Early next week, Debby is expected to get caught up in the jet stream and die in the North Atlantic.

Why the 2am run of the GFDL failed
The 2am EDT run of the GFDL model failed on 97L this morning, so I was quoting the results from the 8pm EDT run last night. I got an email from Morris Bender of the GFDL project this morning on why the 2am GFDL run failed:

The vortex initialized from our initialization was very weak as the initialization process spins up the storm to match the observed initial winds. Since the easterlies in the lower level were very strong in the GFS analysis, it initialized a disturbance with very weak vorticity.

As you can see at hour 0 there was almost no circulation initially. As a result, our grid movement shut down at 3 hours and the inner nests could no longer follow the storm for the rest of the forecast.

The previous runs also had a very weak initial disturbance but there was enough of a pressure gradient that the inner nests continued to follow the vortex and eventually with the high resolution it developed into a significant tropical cyclone.

In the 6z run, the weak disturbance moved out of the high resolution inner grid, into the coarse resolution, and so all we had left was a very weak disturbance that could not be resolved in the coarse outer mesh.


As you can see, getting the computer models to work on weak disturbances is a difficult business! We should not put too much faith in the computer models for any weak system; it is too difficult for the models to get the starting conditions of the storm correct.

I'll have an update this afternoon after the Hurricane Hunters check out 97L.

Jeff Masters