If you’re not aware, there is a hurricane in the Caribbean basin named Sandy and it’s currently moving through the island nation of Jamaica. It will likely dump several inches of rain, causing flooding in many areas of the island. Sandy is forecast to move northward through eastern Cuba and the Bahamas. It is forecast to weaken into a strong tropical storm as it passes the Bahamas.
After that… well, let’s just say that’s where the “fun” starts.
By “fun”, I mean there is the potential for a very large and destructive storm to hit the mid-Atlantic and/or Northeastern US. The “fun” isn’t in the destruction – it’s in the model variability that has existed for the past few days. Let me take you back to Monday when the GFS (American model) forecast Sandy to go out to sea after moving through the Bahamas.
You can see Sandy harmlessly heading out into the central Atlantic. At the same time, the European model had this forecast for the morning of Tuesday, October 30:
You can see the storm is much further north and west on the European model. It’s also much stronger (central pressure of 944 mb on the European model; 968 mb on the GFS). Still, we’re talking about something that is 192 hours away, so a lot can change. For the record, the forecast track of Sandy from the National Hurricane Center reflected the GFS solution.
Here are the model forecasts for the same time (morning of Tuesday, October 30) from today’s midday runs of the GFS (first) and European (second) models:
Both models have shifted further northwest, with the European model now having Sandy over eastern PA. I didn’t show it here, but the GFS model actually pulls Sandy almost due west into Lake Ontario Wednesday morning. So, the GFS still bring the effects of Sandy into the Northeast. Furthermore, as recently as yesterday, the GFS ensemble tracks were split between Sandy heading out to sea and turning back into the US. Here’s what the ensemble tracks look like from the midday (12Z) run:
All but three of the 20 members turn northwest into New England or the northern mid-Atlantic. (Ignore the thick, black line representing the “mean” track as it’s getting tugged by the three outliers.) This is strong agreement amongst the ensemble members. Combine this with the GFS adjusting closer to the European solution (not to mention the Canadian and Navy models, which are similar to the European) and it’s becoming more and more likely that a major storm will hit the Northeast sometime early next week.
What we don’t know is the exact track this storm will take, how much precipitation will fall, what type of precipitation will fall where, and exactly how strong the winds will be. We also don’t know whether it will still be called Sandy at that point as it may have lost its’ tropical characteristics. This doesn’t really matter, of course, as it’s the impacts on society that count the most. I’m just worried it might cause confusion if 24 hours from landfall we stop calling it Sandy. (And if The Weather Channel then decides to give it a Winter Storm name, I might completely lose it.)
We’re a long way from honing in on the exact impacts of this storm, but those along the East Coast need to begin preparing for a potentially significant event.
(One caveat: this type of track is extremely rare. Most storms that track up the East Coast move off to the northeast and continue out into the northern Atlantic. Although it is not unheard of, it is rare that a storm moving up the coast gets pulled back westward into the US. For that reason, I’m still a little skeptical of these model solutions. However, the one outlier – the GFS – has changed to agree with the other models. So, besides climatology, there’s not much you can point to and suggest this will just go out to sea.)