Last night, the GFS simulated the development of a tropical storm over the southern Caribbean. It intensifies the storm to around hurricane status, carrying it northward to a position just east of Florida and then off the coast of North Carolina.
Before getting alarmed, this simulation is more than two full weeks into the future, when the model has no skill. Sometimes we call this model fantasy-land. And for the record, the next run of the model early this morning loses the storm (although some of the model ensemble members show it).
So do we make anything of last night’s simulation?
The blog CrownWeather has a detailed look at this system and isn’t dismissing the idea of some very early season tropical development, stating there is some support for the idea from other models.
Furthermore, CrownWeather notes there is a precedent for occasional tropical development in late May:
The overall weather pattern that is forecast to set up for the second half of May is one that can lead to early season tropical development in either the western Caribbean or somewhere near the Bahamas. Many of the analog years that I have been looking at for this upcoming hurricane season have had early season tropical development. These years include 1951, 1957 and 1976; in fact, 1951 and 1976 had tropical development in May with Category 3 Hurricane Able developing just off of the US Southeast Coast in Mid-May of 1951.
Since 1870, there have been 32 off-season tropical storms (or hurricanes) in the Atlantic (or Gulf of Mexico), 18 of which occurred in May (Wikipedia).
I queried Dr. Greg Postel, CWG’s tropical weather expert for his thoughts on storm potential:
The pattern that may show up in the next one to weeks is one we’d look for in late summer or early fall for tropical development. History has shown us it’s not too early. There are pieces to puzzle that suggest it’s not an unreasonable idea. But an operational run of tropical development at 300+ hours is just unreliable.
Postel further noted the development models are showing is not strictly tropical, but intricately involved with interactions with mid-latitude weather.
If nothing else, what the models are showing is a sign hurricane season is just around the corner. We’ll let you know if this idea of some early season development gains momentum...