GFS Model Pulls Left But Can It Be Trusted? Waiting for the Euro (Again!)

I mentioned on my post about Hurricane Joaquin at 11pm that when you look at loops and current weather going on it was making me a little cautions. The NAM model is still holding on to the idea that the upper trough in the Southeast US captures the hurricane and drags it inland near Myrtle Beach! This was similar to the Canadian model idea today but it was discounted by the fact the European continued its eastward plan for Hurricane Joaquin.

Tonight’s GFS model went decidedly left (again) and now comes perilously close to the coastline. It is still and offshore miss for the northeast though it does clip Southeastern New England.

gfs60 gfs72

Here we have the GFS model 60 and 72 hour forecast which takes us through Sunday. From here the hurricane moves north northeast to very close to 40N & 70W. Let’s look at the tonight’s 96 hr position which is Monday evening verses the 102 hr position of the prior run.

gfs96 gfs102

That is quite a shift westnorthwestward. What the model is doing is that it is showing that the hurricane is pulled back westward by the upper trough early in the game which puts it a good 300 miles further west than the prior run.

gfs78

After this time frame the hurricane is pretty much on its own as it moves way and the whole trough moves offshore. Now let’s look at the Nam model  for the same time frame.

nam78

The NAM has no indication of a separate feature because the upper low in the south grabs the whole thing and becomes one. This is a key difference and it is why the nam was so far left. The hurricane in the NAM model does not maintain a separate identity where as on the GFS and on the European they remain separate. The difference in the European and the GFS model is that the European has the upper low acting like a football team punter basically kicking the hurricane away to the northeast.

euro96

So now what do we do? Well we certainly want to be cautious given the GFS look tonight. These are 3 decidedly different outcomes for three different reasons. 3 models have 3 different ways of handling this dilemma. From a forecasting standpoint we are obviously going to wait for the European. I am going to assume it is going to probably do the same thing it has been doing. However if by some chance it comes in west with a different idea then all bets may be off and we will have to continue to wait to see which model ultimately wins this battle. We could look at the Canadian and it is only out to 54 hours. Guess what? Another solution.

gem60 gem54

Here we have the upper low pulling the hurricane very tightly into its flow and even further left then the GFS. Unfortunately I can’t see beyond this point right now but my bet is it moves it inland in North Carolina which is where this model has been for the last many runs.

I guess all we can do now is wait since I have to get some sleep.

Be sure to download my app and subscribe to my local forecasts available for New York, New Jersey, Connecticut, and Southeast Pennsylvania. The app is free and a subscription is just 99 cents a month. The app is free of advertisements and there are no tracking or security issues.