Most Read: Local

Posted at 11:25 AM ET, 10/22/2012

Tropical storm Sandy likely on the way, possible East Coast encounter next week

There are two areas of disturbed weather across the Atlantic basin: one in the southwest Caribbean, and one in the middle of the ocean far from land. While both disturbances are expected to intensify, the one in the Caribbean - likely to be named Sandy - could potentially affect the East Coast of the U.S. next week.

Enhanced satellite image of tropical depression 18 south of Jamaica. The approximate location of the surface Low is marked with a red “L”. (NOAA)
This Caribbean disturbance, about 300 miles south of Jamaica, was classified as tropical depression 18 this morning. Its minimum pressure is 29.62” (1003mb) and is forecast to gradually strengthen before making a turn toward the north later today into tomorrow. That would take it over Jamaica, then central-eastern Cuba, then the Bahamas.

Regardless of the intensity (in terms of peak wind speed), these areas need be on high alert for heavy rain and the associated flash flooding and mudslides. An aircraft will be flying into it this afternoon to more accurately determine its organization and intensity.

The key question becomes, where does it go after it passes the Bahamas?

When it comes to predicting this disturbance’s future, there are many types of models that are routinely run. Some cover the entire globe and are run every 6 or 12 hours (such as GFS and ECMWF), while others are more regional in scope with higher resolution (such as NAM, RAP), and others have even higher resolution and are designed specifically for hurricanes (such as HWRF and GFDL) and are only run on demand when there is an active disturbance or storm.

The typical run from each of these models is called a deterministic run, while some models also produce an ensemble run. In a typical ensemble, the initial condition (state of the atmosphere based on observations that tell the model how to start) is varied slightly to effectively produce many possible scenarios. This helps to account for uncertainties in the observations as well as inherent errors in the model. It also allows for a more probabilistic forecast as opposed to the single solution provided by a deterministic forecast. You could also create a multi-model ensemble from all of the various deterministic runs.

Various model track forecasts for the disturbance near Jamaica. (SFWMD)
In the plot shown here, track forecasts from several models are included (you can ignore the extrapolated, climatological, and basic ones: XTRP, CLP5, BAMD, BAMM, BAMS). The ones I want to specifically point out are the blue one with diamonds (AVNO), the purple one with squares (AEMN), and light gray ones without symbols (APxx). AVNO is the identifier for the deterministic GFS run, AP01-AP20 are the twenty members of the GFS ensemble, and AEMN is the ensemble mean of those twenty members. You can see the added value of an ensemble... many more realistic possibilities come into view.

While the majority of models keep this system well offshore, there are a small handful that nudge it further west and bring it into the Northeast U.S. next Tuesday.

Surface pressure, thickness, and precipitation from one of the twenty GFS ensemble members valid next Tuesday. (Univ at Albany)
The one (from a GFS ensemble member) that happens to curve left into Long Island is certainly ominous, but shouldn’t be taken too seriously: it’s eight days away from landfall and that solution is an outlier. The surface pressure and precipitation map from that particular ensemble member valid next Tuesday is shown here for reference. Given the potential impact though, it’s worth some attention.

European model simulation of surface pressure, thickness, and precipitation Monday night. (
Another leading track model is the ECMWF (not shown above, but pictured to the right). It brings this storm to the mid-Atlantic coast next week... as a hurricane or possibly a hybrid tropical/Nor’easter storm.It simulates flooding rain, strong wind, and destructive storm surge and coastal erosion affecting every state in the mid-Atlantic and Northeast. The model also would suggest snow at higher elevations in the interior. This simulation is also a bit of an outlier, but since the ECWMF is historically one of the better models, it should not be dismissed.

While a massive landstrike is less likely than an out to sea scenario, the entire U.S. East Coast from Florida to Maine should at least be casually aware of this storm scenario, especially those who are immediately on the coastline. Always keep in mind that storm conditions may extend hundreds of miles away from the centerline of the track, and dangerous surf and coastal flooding/erosion are possible with the majority of these tracks.

Elsewhere, there’s another disturbance located approximately 700 miles east-northeast of the Leeward Islands. While it is fairly close to becoming a tropical depression, its window of opportunity is closing. The vertical shear is a reasonably low 10 knots today, but is forecast to increase dramatically by Thursday. Though it may intensify to a tropical storm and earn a name - Tony, it won’t be getting close to land. It will move north for another day or so, then accelerate to the northeast as it gets absorbed by a mid-latitude trough.

* Brian McNoldy is a senior researcher at the University of Miami’s Rosenstiel School of Marine and Atmospheric Science.

By Brian McNoldy  |  11:25 AM ET, 10/22/2012

Categories:  Latest, Tropical Weather

Read what others are saying

    © 2011 The Washington Post Company