More Atlantic Good News!

The National Hurricane Center says nothing in the way of a tropical storm for the next seven days. Yay!!

Looking a little farther out, the single most accurate model for hurricane forecasting, the ECMWF, says nothing for the next ten days, which takes us to August 31 (below). The ECMWF now has an adjunct version that uses artificial intelligence; it says no storms, also. Important note: this is the first hurricane season for AI adjunct so we have no idea how skillful it may be.  

None of the other models that reliably forecast tropical weather are forecasting any tropical storms through ten days. In fact, the model consensus until the end of the month is for abnormally calm conditions. 

Addition: Matt Lanza, a Houston-based meteorologist who is a fine hurricane forecaster, published a piece on this same topic today. It begins:

Coming into this hurricane season, there were breathless forecasts calling for anywhere from 20 to even 30 named storms from some folks. Government forecasts were more conservative but even they were the most active we’ve ever seen from NOAA. Colorado State, too. So, as we sit here on August 21st with next to nothing showing up in modeling for the next 7 to 10 days, it’s reasonable for the general public to ask questions about those forecasts.

I highly recommend Matt's piece for those interested in the meteorology as to why we've had a low number of tropical storms so far this season. However, I disagree with this:

So despite the perception that this season has been slow to start, it actually has not been. In fact, 2024 is in rarefied air in terms of ACE to date. This perception might have to do with storm inflation, or the idea that we are naming more storms today than we did, say, 20 or 30 years ago. In 2020, basically the benchmark for recent historically active seasons, today marked the formation date for Hurricane Laura, the 12th named storm of the season. Here we are in 2024 sitting at five, and it’s no wonder the perception is skewed. Despite having more than twice as many storms to this point in 2020, we only had half the ACE (26 vs. 55 this year). 2020 kept churning out mostly sloppy storms until Laura, and then things got nasty. We’ve had three legitimate hurricanes this year and only about one “throwaway” storm (Chris). 

I completely agree with Matt about the Hurricane Center's name inflation (Chris never would have been named 20 years ago) but that is a topic for another time.

Here is why I disagree with bringing the ACE into the discussion. Assume you went a Vegas sports book and forecast the Yankees would win 100-110 games this season. If they had only won 55 games at this point, you'd have every reason to worry about your bet. But, if you went to the cage and said, "I'd like to collect on my bet because they have hit a lot of home runs," they'd laughing while they threw you out the front door onto the Strip. 

While number of home runs may be related to number of games won, they are two different metrics. If, let's say, there are 11 named storms then the forecasts are busts regardless of how high the ACE may or may not be.

I completely agree with Matt that September and October might be extremely active. But, I'm cheering for the forecasts in the first paragraph to be wrong as a bad hurricane season causes tremendous economic losses, dislocation and puts thousands of lives in danger.