Hugging the storm

The usual suspects seem to be hyping Invest area AL95. TWC has it on their front page (“Possible New Threat Emerges, Here’s Where It’s Headed”) and some Florida sources are HYPErventilating (WESH in Orlando: “Models show Invest 95-L may near the southeast US coast this week”). So should you worry? No. Enjoy your labor day, get your week started, check back Wednesday (Caribbean) or Friday (US). If you insist on more than that, read on …

NHC has had a disturbance coming off the coast of Africa on their outlook for several days now, and assigned it the temporary ID AL95 to start model tracking. As of this morning (4 Sept 2023) at 2am they show the chances of development at 60% in the next two days, and 90% in the next week. There is good reason for their enthusiasm, the major global models have been fairly consistently showing potential development in that area towards the middle or end of this week. Let’s start with the tracks – remarkably consistent for something that isn’t much more than a cluster of clouds …

click any image to embiggen.

I tend to be a bit suspicious at tight groupings at this stage – it could mean they are in agreement and it’s a clear picture, and that is the more likely answer, but it could be an initialization issue and they are “consistently wrong.” We will just have to watch to see when and where the storm actually forms, and what the surrounding conditions (especially the high pressure to the north that will ultimately “steer” the storm) do. It’s a long way out there, but the most likely scenarios show the system skirting the Leeward Islands and heading north offshore from the US. HOWEVER, that’s at best pure speculation at this point – saying things like “may near the southeast US coast this week” are just shoveling out clickbait, and anyone doing so is a bad person and should feel bad 😛 . Sure, it’s what some models are showing in ten days, and yes we’ve gotten better at such things, but would you trust a ten day old weather forecast? Didn’t think so. It’s “interesting” but just not “actionable,” much less worth getting excited over.

In one sense run-to-run consistency has been pretty good. Compare the 18Z and 00Z runs (latest two available as I write this) forecasts for the 14th:

GFS model run comparison

That looks pretty consist…. wait, what the heck happened in the middle of the Atlantic? A whole second storm appeared and disappeared! So … caveat emptor.

That’s the “where.” Here is what the GFS model primary run is showing in terms of development for the next 5 days …

Notice the next 36 hours the storm isn’t spinning up, but then enters more favorable circumstances and begins consistent growth. Looking at that among the models we get this plot of intensity vs. time …

So they do look like they are all spinning up something. But it’s way to early to say anything actionable.

To emphasize, the official NHC outlook does not have any of the “magic words” that should trigger you to even monitor this system in the Caribbean, much less the US (although I expect the “interests in the Leeward Islands should monitor” to appear by Wednesday). So unless you are a tropical weather enthusiast you aren’t gaining anything but unnecessary stress wearing out your F5 key on this yet.

PS – Hugging the Storm is a song by Grai (youtube link).


  1. Thank you for trustworthy and common sense information! You are greatly appreciated!

  2. Not only do you really know your meteorology, but you’re really funny too! We’re rebuilding a home destroyed in SANIBEL by Ian and we’ve just moved into the Burnside area here in Savannah. Dodging two bullets this season till we can sell SANIBEL can be a challenge. Thanks for helping to keep us “sane”

Leave a Reply