Sunday, October 21, 2012

East Coast May Be Threatened By 'Sandy'

Models are going crazy over the potential for both a snowstorm and a hurricane to hit the northeastern quadrant of the US (Ohio Valley, Northeast, Mid-Atlantic). Let's jump right into this.

First, I will discuss the tropical aspect of this system. Then, I will hone in on the snowfall aspect.

Section 1: TROPICAL

This is the little invest that could cause a very extreme Halloween on the weather spectrum. Invest 99 currently resides in the Caribbean, with a 70% chance of development in the next 48 hours as declared by the National Hurricane Center. The system is expected to stay stationary before moving north. During this time, strengthening into a named tropical system may very well occur, and models are favoring such an idea. Here's a gathering of the models for 48 hours out. If the invest strengthens to a named storm, it will be called 'Sandy'.

All the models are in pretty good agreement that the system will likely strengthen into a tropical storm or something of similar strength. At this point, we can pretty much say that the next 48 hours will not threaten the US mainland.

But there's a different story later on...

The models begin to change forecasts and vary on strengths past the 5 day mark. Let's look at a gathering of the models as far out as they can go. TIP: To see the number of hours this forecast goes out to, look for a 'T = (number of hours) h' on top of each image.

NOTE: Each model's forecast is at a DIFFERENT TIME. See the "T= (number of hours) h" on
the top of the image to see when each model forecast is valid for.
Let's do a model-by-model breakdown of each forecast. Note, confidence levels do not add up to 100% in the end.

This model keeps Sandy modest and only moderately strong as it gets dangerously close to Florida. While this seems the most realistic in comparison to stronger forecasts, the NOGAPS does not have a good track record. Confidence level: 15%

European Model (ECMWF):
The ECMWF is an amazing model and commonly claims the best forecasting record when it goes up against other models. It is showing a very strong landfall in the Northeast/Mid-Atlantic, which could seriously hurt the cities closest to the coastline. Despite the ECMWF's great track record, this is the long range forecast, and that must be factored in to my confidence level. Confidence level: 40%

European Ensembles (ECMWF Ensembles):
These are the ensembles of the ECMWF model that combine into an amazing forecast source. Based on the fact that these ensembles originate from the ECMWF model, and that there are so many (I heard 52 members from somewhere) of the individual members in the ensemble set, my confidence level is pretty high on this one. Confidence level: 55%

Canadian Model (CMC/GGEM):
The Canadian model is horrible at long range forecasting, and is absolutely terrible at tropical cyclone forecasting. I'm not even going to think about this model's forecast for the time being. However, the fact that it holds some similarities to the ECMWF and GFS is slightly encouraging. Confidence level: 5%

American Model (GFS):
The GFS claims second place in the best model race right behind the ECMWF. A recent update to the GFS has drastically improved its forecasting abilities, threatening the ECMWF for the title of best model. The ECMWF and GFS are fairly similar, but the ECMWF goes inland while the GFS does a 'drive by' on the East Coast, where it hits the region and goes back out to sea. I'll need some more runs of the GFS to be more confident in this one. Confidence level: 30%

All in all, I believe the East Coast should definitely be prepared. It's always better safe than sorry, and the fact that so many models are even thinking of threatening the East Coast is worrying. Watch for more information coming in the next few days.

Section 2: SNOWFALL

This is the part many of you have been waiting for. So, without any further ado, let's check out the snowfall forecasts.

NOTE: This is material from AccuWeather Pro and should be treated as such;
if you are going to use this image, please reference Accuweather Pro.
This is the ECMWF snowfall forecast for 240 hours, of 10 days away. As you can see, the effect on Sandy going inland is huge- over 20 inches of snow hits Michigan and even more in southern Canada! Several inches of snow are also found in Mid-Atlantic states and parts of the Ohio Valley and Northeast.

This is indeed big, as it is from the ECMWF model. However, one must remember that there are many models out there, each with their own possibilities. So, take the above for what you think it's worth (I don't think it's worth much at the moment), and let's move on to the GFS.

The GFS is much less enthusiastic, and it should be. This forecast of the GFS has the hurricane get dangerously close to the coast before turning back and heading off into the ocean. Again, it's much different than the ECMWF, as the latter goes inland into Canada. Also, both models are in the long range, and all long range forecasts don't have the best chances of verifying, no matter the model.

It will depend on where the system ends up to determine just what (if any) snow falls on the eastern part of the nation. One slight movement in any direction could drastically alter Sandy's course and thus the effects it will have.



Anonymous said...

It would be great if Michigan got snow but I have trouble trusting anything that comes from Accuweather.

Andrew said...

The data is from the ECMWF, it's that AccuWeather is the one that provides this image.

Logan said...

I agree with anonymous, I don't like accuweather :-)

Anyway, nice informative post :)

I'm just hoping for snow, whether it be an inch or a flurry, I don't care just some snow from this in Louisville

ERN WX said...

An ECMWF/GFS commpromise is hilariously what I am forecasting now, and expect will change. I see that the GFS is tooooo strong with its trough. This storm will become a noreaster, when it reaches cooler waters. I am ready for a cat. 5!!!!!!! Don't worry a cat 5 won't happen!!!!!!!!

Anonymous said...

We can always count on the models making things exciting. Will it snow? Will there be a hurricane hitting the east coast? Will nothing happen?! We kinda see all those scenarios going on in the models. Gotta love it.

Cam Smith said...

Do you think NJ will get some flurries soon?

Anonymous said...

if the track is changed could philly get 5 inches of snow or more