WunderBlog Archive » Category 6™

Category 6 has moved! See the latest from Dr. Jeff Masters and Bob Henson here.

Little change to 93L; Brazilian floods kill 42

By: Dr. Jeff Masters, 2:18 PM GMT on June 23, 2010

There has been little change over the past 24 hours to the tropical wave (Invest 93L) located a few hundred miles south of Haiti. The storm has brought heavy rains to the waters south of Hispaniola, with radar precipitation estimates from the Puerto Rico radar of over six inches of rain in the past day. The heaviest rains have avoided land areas so far. Satellite loops show a very disorganized system, with no low-level spiral bands and only limited heavy thunderstorm activity. There are no signs of a surface circulation visible on satellite imagery. Pressures at the ground stations surrounding the storm (Bahahona and Kingston) are not falling. Water vapor satellite loops show that moist air surrounds 93L, though there is a modest patch of dry air to the storm's southwest. This dry air is likely contributing to the lack of heavy thunderstorms on 93L's west side, slowing development. Wind shear is a low 5 - 15 knots. The high wind shear associated with the strong winds of the subtropical jet stream are over the northern Caribbean, too far north to interfere with development, but close enough to provide good upper-level outflow for the storm. Visible satellite loops show high level cirrus clouds streaming away from 93L to the northeast, evidence of the upper-level outflow channel that is developing to the storm's north. Sea Surface Temperatures are plenty warm, a record 29 - 30°C. The Madden-Julian oscillation (MJO) currently favors upward motion over the Caribbean, which will act to increase the chances of tropical storm formation this week. The main negative for 93L continues to be the lack of spin. The University of Wisconsin 850 mb relative vorticity analysis is showing that spin at 850 mb (roughly 5,000 feet in altitude) has increased over the past day, but 93L needs to acquire additional spin before it can grow more organized. The Hurricane Hunters are on call to investigate 93L Thursday afternoon. Today's flight was canceled, due to 93L's lack of development.

Figure 1. Morning visible satellite image of the central Caribbean disturbance 93L.

Track forecast for 93L
NHC is giving 93L a 30% chance of developing into a tropical depression by Friday morning, which is a reasonable forecast. Given the storm's current lack of spin and relatively modest amount of heavy thunderstorms, the earliest I'd expect 93L to become a tropical depression would be Thursday afternoon, with Friday more likely. Wind shear is expected to be low, less than 10 knots, over the central and western Caribbean this week. Water temperatures will be warm, dry air limited, and the MJO favorable. I don't see any major impediments to the storm becoming a tropical depression by Friday, except for possible interaction with land. There is a low (less than 20% chance) of 93L becoming a hurricane in the Caribbean. Expect 93L to bring flooding rains of 3 - 6 inches to Jamaica, eastern Cuba, and southwestern Haiti today through Thursday. These rains will spread to the Cayman Islands and central Cuba by Thursday through Friday, and western Cuba and Mexico's Yucatan Peninsula on Friday and Saturday. The current run of the SHIPS model has 93L slowing down late this week to a forward speed of just 7 knots (8 mph) from its current speed of about 10 mph, in response to a weakening in the steering currents. A trough of low pressure is expected to swing down over the Eastern U.S. on Monday. If this trough is strong enough and 93L develops significantly, the storm could get pulled northwards and make landfall along the northern Gulf of Mexico coast. This is the solution of the ECMWF, GFDL, and HWRF models. This seems like a likely solution, since recent runs of most of the models have been showing a stronger trough. If 93L stays weak and/or the trough is not so strong, the storm would get pushed west-northwestwards towards the Texas coast. This is the solution of the NOGAPS and Canadian models, which foresee a more westerly track for 93L across the Yucatan Peninsula, with the storm eventually threatening Texas by Monday or Tuesday. The GFS and UKMET models do not develop 93L. A likely landfall location is highly uncertain this point, and the storm could hit virtually anywhere along the Gulf of Mexico coast from South Texas to the Florida Panhandle given the current uncertainty in its development, and the strength of next Monday's trough that may steer 93L northwards.

Intensity forecast for 93L
Forecasting the amount of wind shear in the Gulf of Mexico next week is problematic. There is currently a band of high shear near 30 knots over the Gulf, and most of the models predict that this band of high shear will lift northwards, keeping low wind shear over the Gulf next week. This should allow 93L to intensify to at least a 50 mph tropical storm, as predicted by the HWRF model. The GFDL and ECMWF models are calling for 93L to become a Category 1 or 2 hurricane, and this is certainly a possibility. I expect 93L will become Tropical Storm Alex in the Gulf of Mexico next week, and give a 20% chance that it will become a hurricane.

Elsewhere in the tropics
None of the reliable computer models is calling for tropical cyclone formation in the Atlantic over the next seven days.

Figure 2. Total accumulated precipitation in millimeters (left) for the 7-day period ending at 8am EDT June 21, 2010. The yellow colors (200+ mm, 8+ inches) are where extreme flooding was observed. Satellite image from NASA's MODIS instrument (right) for June 17, 2010, showing heavy thunderstorms over the Alagoas state of Brazil. Image credit: NASA and NOAA Climate Prediction Service.

Floods in Brazil kill dozens
Heavy rains over the past week in the northeast state of Alagoas in Brazil have led to major flooding that has killed at least 42 people. Six hundred people are still missing. This weekend rains in Alagoas and Pernambuco states are the latest in a series of devastating floods to strike Brazil this year. Since the start of Brazil's rainy season last November, 488 people have been killed by flooding and 7.5 million have been affected in 10 states. Much of the heavy rains can be blamed on El Niño, which ended in May. In April, flooding and landslides triggered by torrential rain killed at least 229 people and did $13.3 billion in damage in the Rio de Janeiro area.

It's been a bad year for floods, and there are two other major flooding disasters that have occurred in the past week. In China, the death toll has risen to 211 people, with 119 people missing, because of flooding in the southern portion of the country. France suffered a flash flood last week that killed 25.

Wind and ocean current forecast for the BP oil disaster
East to southeast winds of 10 - 15 knots will blow in the northern Gulf of Mexico today through Sunday, according to the latest marine forecast from NOAA. The resulting weak ocean currents should push the oil to the west and northwest onto portions of the Louisiana and Alabama coasts, according to the latest trajectory forecasts from NOAA and the State of Louisiana. I would expect Mississippi to have its most serious threat of oil yet early next week as these winds continue. The longer range outlook is uncertain, and will depend upon what 93L does.

Resources for the BP oil disaster
My post, What a hurricane would do the Deepwater Horizon oil spill
My post on the Southwest Florida "Forbidden Zone" where surface oil will rarely go
My post on what oil might do to a hurricane
NOAA's interactive mapping tool allows one to overlay wind and ocean current forecasts, oil locations, etc.
Gulf Oil Blog from the UGA Department of Marine Sciences
Oil Spill Academic Task Force
University of South Florida Ocean Circulation Group oil spill forecasts
ROFFS Deepwater Horizon page
Synthetic Aperture Radar (SAR) imagery from the University of Miami

Jeff Masters

Flood Hurricane

The views of the author are his/her own and do not necessarily represent the position of The Weather Company or its parent, IBM.