Cause for concern over possibility of a tropical cyclone in the Gulf of Mexico increases
June 22, 2010
As I noted in my morning post, the NHC’s assigned probability of the tropical wave in the Caribbean, Invest 93L, developing in the next 48 hours dropped to 20%, which is where it remains in tonight’s Tropical Weather Outlook. The cause for the drop is not exactly clear. While one can easily state the reasons it’s not higher (lack of an obvious circulation at the surface, climatological improbability, etc.) the presence of those same factors 24-36 hours ago makes one wonder how the probability got so high in the first place. Anyhow, here’s how the Caribbean looked before sunset this evening:
The presumed center (scientific dart throw) was placed at 15.6N 75.2 W, which is the southwest edge of the cloud mass that is southeast of Jamaica. In this image, one can see the clouds associated with heavy rain that is streaming into the Dominican Republic and Haiti. The Tropical Weather Discussion mentions that 4-7 inches of rain is expected on the island. (in a post on the previous invest, 92L, Dr. Masters opined that 4 inches was “probably the lower threshold for life-threatening floods to occur in the Haiti earthquake zone”). Looking to the Pacific, at the extreme left edge, one can see the outskirts of Hurricane Celia, with what will be Tropical Depression Five to the east.
At the end of my post last night, I framed 93L as a fight against climatology. The image below shows all of the tropical cylones that have formed in the month of June since 1950 (courtesy of Hurricane Historical Tracks)
As you can see, past history shows that it is almost unheard of for a storm to form in the eastern Caribbean this time of year. This area is often referred to as “the graveyard” or the “cyclogenesis hole” alluding to how seemingly strong systems die out or otherwise fail to develop in this area. However, 93L has cleared the graveyard and is nearing the point where storms have tended to form.
The global models that I trust to forecast cyclogenesis aren’t showing development in a reliable forecast period (<7 days). However, for the past two runs, the specialized GFDL model has shown a tropical cyclone developing as 93L approaches the Yucatan. The 18Z run brings it near major hurricane status at the end of the 5 day forecast period, positioned 160 miles to the southwest of the Deepwater Horizon. Additional cause for concern comes from the similarly specialized HWRF. While runs up through this morning did not show clear-cut development, its past two runs do show a weak tropical storm forming and going into the Gulf of Mexico, as well (18Z run positions it ~180 miles south of the Horizon). The final bit of concern comes from forecast low-level flow. It had looked like that if the wave didn’t develop, it would proceed squarely into the Yucatan Peninsula, thereby minimizing any further threat of development. However, it is starting to seem that even without forming into a storm, the wave would tend to go northwest, brushing the tip of the peninsula, thereby continuing the threat.
The challenge (if not impossibility) of determining a likely path for a storm, if one were to develop, is knowing where it would form. Forecast errors associated with a small error (10 mile) in positioning a storm center multiply quite rapidly as the forecast period lengthens. With the uncertainty in the position of a (possible!) circulation being on the order of 100 miles, the resulting forecast errors would be extremely large.
A hurricane hunter flight is penciled in for tomorrow; if it appears there’s nothing to examine, the flight can be canceled. Because of the increased interest in the possibility of a tropical cyclone in the Gulf, the probability of cancellation may be less than normal. Even with that in consideration, I think it’s it likely that the mission, scheduled to launch at 9:30 tomorrow morning, gets pushed to the right 24 hours.