Model silly season may begin soon in the tropics

Welcome back to The Eyewall! We’re going to fire back up some regular updates now as hurricane season approaches. The plan is currently to post an update each Monday, Wednesday, and Saturday until activity necessitates more frequent posts.

Headlines

• GFS model periodically trying to spin something up around Memorial Day in the SW Atlantic or Caribbean.
• While that may or may not happen, it’s about this time of year that the GFS model tends to do some wild things that are usually unrealistic so use it carefully.
• Eastern Pacific may kick up some activity later next week.

Checking in on the tropics: A reminder about the GFS model

Over the last week or so, we’ve started to see some signs in weather models of some very late period mischief in the Caribbean. This is especially true on the GFS operational model. For those of you that track these things, it’s around this time of year that the GFS begins to go a little haywire with tropical activity traditionally. For whatever reason, it likes to latch onto potential tropical disturbances and blow them up into well-organized storms, usually in the Gulf or Caribbean in May and early June. As most of you know, the chances of a very well-organized May or early June storm is extremely low. So we always tell people to generally put little stock in the GFS operational model this time of year in the tropics.

Remember, operational models are known as “deterministic” models. They’re one run, one solution. As meteorologists, we have learned to use an ensemble approach to forecasting, particularly when trying to determine if tropical development is realistic in a medium-range to extended range timeframe. Ensembles are where they run these models anywhere from 30 to 50 or so different times. Each time, something is tweaked in the initialization (what the model is seeing at hour 0), and then the model is run out 240 to 384 hours. This allows us to capture a more realistic envelope of outcomes from the modeling that informs our thoughts on development, risks, etc.

Anyway, the thought process right now is that conditions may become slightly favorable in the Caribbean or Southwest Atlantic for something in about 10 to 15 days. But I would encourage a lot of caution about getting worked up regarding some of the GFS model runs. They may begin to show some outcomes that are unrealistic.

In the meantime, keep an eye on the Eastern Pacific. It may be time to fire up that basin in about seven to ten days. We’ll take another look at things on Wednesday.