Matt Lanza

Thanks to Eric for covering yesterday. Much like we do with Space City Weather, he and I will split duties, though I’ll spearhead most of this site, as he does with Space City Weather. Today we’ll discuss more about why the GFS model is probably getting the day 10+ range wrong. We’ll also tackle part 3 of our seasonal outlook which discusses water temperatures in the Atlantic.

One-sentence summary

The tropics are quiet, and we continue to believe that what is shown on the GFS model late in its forecast period is not a serious concern.

Happening now: Not too much!

All is quiet across the tropics this morning with no disturbances of note. The main weather feature right now is an upper low near New England that has been responsible for directing Canadian wildfire smoke into the East.

While this is not directly related to tropical weather, it was noteworthy to watch the squashing of fair weather clouds as smoke moved into some locations in the Northeast. Reminiscent of how Saharan dust can help inhibit tropical systems. We’ll touch more on the dust situation next week.

The medium range (days 6-10): Nothing yet

We continue to see quiet conditions likely to continue through about day 10, so there is not much to discuss here.

Fantasyland (beyond day 10): The GFS is probably still wrong

We’ve talked about this for a few days now, but we also know that when something shows something adverse consistently, it can spark some concerns. Despite the fact that the GFS operational model has shown a substantial storm for multiple straight runs in the day 12 to 16 timeframe, we remain fairly convicted that it’s really nothing to worry over. Besides the fact that the GFS frequently gravitates toward “scareicanes” in the Gulf this time of year, Eric also touched on the ensembles not supporting it yesterday. That remains true today.

Operational model runs are really just that: One run of one model. It’s deterministic. This is what the model thinks is going to happen, and that’s that. Ensemble guidance is much more nuanced. You take those models, like the GFS or ECMWF (the Euro) and run it 30 to 50 different times with tweaks to how the model is initialized, the starting point that it bases the rest of the forecast on. By doing this, you get a more realistic set of possible outcomes than just what one deterministic model shows. And in the case of this rogue GFS model storm, the ensemble guidance offers little additional support for something like the operational model shows. And the European model? It shows nothing at all.

That said, there’s good reason to think that something may be afoot late. I touched on this a little back on Monday. But the overall background pattern is going to become a bit more supportive for rising air in the Atlantic.

Couple that with an upper level pattern that shows a mean trough (dip in the jet stream) over the Southeast, eastern Gulf, and western Caribbean, and that argues that, broadly, we look more favorable for tropical development toward the final 7 to 10 days of June. But probably not like what the GFS operational model is showing at the moment. We’ll see if this can roll forward as we move into next week.

The 2023 seasonal outlook Part 3: Atlantic Ocean water temperatures

On Tuesday we discussed what the outlooks were for the 2023 Atlantic hurricane season. Yesterday we broke down El Niño. Today, let’s look at the Atlantic Ocean. With the exception of the northeast Gulf of Mexico, the area off the immediate Southeast and Mid Atlantic coasts, and a chunk of the subtropical Atlantic, the entire basin is running hot, with widespread water temperature anomalies that are above normal.

As Michael Lowry pointed out in his (always excellent) daily tropical newsletter yesterday, the eastern portion of the Atlantic is at its warmest levels on record for this time of year, and it is not really even close. Obviously, this is important. But just how important? If you look at page 9 of Colorado State University’s seasonal hurricane outlook from this month, you’ll find that eastern Atlantic SSTs have a nearly 0.6 correlation to seasonal accumulated cyclone energy (ACE). In simple terms: A warm eastern Atlantic this time of year typically produces active hurricane seasons.

It’s tough to find a true analog to this year though because a.) this is the record warmest and b.) no previous upper-end warm eastern Atlantic basin temperatures coincided with an El Niño event in the Pacific. But, looking at the Atlantic in a box, if you took the five warmest years outlined in CSU’s forecast behind 2023, you averaged 23 named storms. Big yikes. Compare this to what we said about El Niño yesterday, which averaged 9 storms. Those goalposts are wide enough for <insert the name of a football kicker who has undoubtedly made you angry at some point> to never miss. No one ever got into weather forecasting because it was easy.

So El Niño implies a weak hurricane season, while the current water temperature situation in the Atlantic implies an extremely active season. This is a tug of war and a half, and one of the most fascinating seasonal forecasting challenges I’ve ever seen. Tomorrow we’ll talk about some risks and tie a ribbon on all this.

Happy Tuesday to all. It’s very early in hurricane season, but at the very least, we’re trying to make sure you have a reason to check us out each day with something interesting. Today we’ll have the first of four parts on the seasonal forecasting challenge for this particular hurricane season. We’ll discuss the forecast today, El Niño tomorrow, water temperatures in the Atlantic on Thursday, and then the risks regarding the season on Friday.

One-sentence summary

No meaningful tropical development is expected over the next week or so, though a system near the Azores Islands has a very low chance to briefly become something before the weekend.

Happening now: Strange system swirls near Azores

The fact that we’re even discussing the Azores in June is pretty ridiculous. Storms simply do not form here this early in the season. There’s an argument to be made that maybe 40 or 50 years ago, in this same scenario, we actually would not be talking about the Azores because the technology to monitor these things has improved so much. So in 2023, we can tag disturbances and even name lower-end systems that would not have had that done “back in the day.”

Regardless, it’s impressive and looking at satellite this morning, it’s evident that we have a system out there, however when you go “under the hood,” this is technically not a tropical or subtropical system at this point.

In a future post, we’ll get into the technicalities of these things; what is subtropical versus tropical versus extratropical or cold core versus warm core and why all these things matter meteorologically. The bottom line on this particular disturbance is that it has about 24 to 36 hour window to develop some warm core (tropical) characteristics, as it is moving over an area with sea surface temperatures that are anywhere from 2°F to 5°F above normal.

That said, the current 10 percent probability offered by the National Hurricane Center seems fair, as the water is still fairly cool for true tropical development here.

The medium range (days 6-10): Still quiet

There’s nothing new to add in this period today, so we’ll keep this simple. All’s quiet.

Fantasyland (beyond day 10): No changes in thinking

Yesterday’s post discussed why things may become more favorable for something later in the month. Nothing has changed today regarding that topic. The GFS operational model continues to spit out a somewhat spurious looking system around days 13-16. It continues to be mostly on its own bringing it northward toward the Gulf, lacking any real meaningful ensemble (multiple runs of the same model with tweaks) support for that at this time. We will get into the operational versus ensemble model differences sometime soon.

The 2023 seasonal outlook Part 1: The forecast

Alright, let’s talk the seasonal forecast. We will break this out into four parts over the coming days so as to not inundate you with too much stuff at once. Today, we’ll explore the forecasts. Tomorrow, we’ll discuss El Niño. On Thursday we’ll look at water temperatures across the Atlantic basin. We’ll wrap things up on Friday by putting this all together and discussing the risks.

The caveat I have to start with: Seasonal forecasts are mainly scientific research exercises. You should prepare each season as if it will be the one where a storm strikes. Below average seasons can be bad ones too. Last year was near average and produced Hurricane Ian. 1992? A pretty quiet season, but it produced a storm named Andrew. 1983? One of the quietest hurricane seasons on record. It produced a storm named Alicia for Houston. So whatever you see with respect to seasonal forecasts, it’s honestly not exactly useful information. Fun information? Sure. Useful? Not so much. Got it? Great.

The consensus for the upcoming season is “near average.” What does that mean? A consortium of AXA XL (a reinsurer), Colorado State University, and the Barcelona Supercomputing Center have been aggregating seasonal hurricane forecasts for a few years now. These forecasts come from the private sector, government agencies, seasonal models, and universities. They have a great site where we can get almost a consensus forecast for the upcoming season, as well as see the outliers (what we refer to as an “ensemble forecast” in meteorology). The consensus for this season (which is two storms deep as of today) is for roughly 16 named storms, 7 hurricanes, and 3 major hurricanes, which is near what an “average” season would produce.

I personally tend to put a bit more weight on Colorado State and NOAA’s outlooks because they’ve been at it for a long time, and they offer a good deal of insight into their processes. I will also speak highly of Weather Tiger, run by Dr. Ryan Truchelut. He also has a tropical newsletter, focused on Florida a lot but no less interesting that you may want to add to your repertoire. In my work in the energy industry, I have also crossed paths with CWG (Commodity Weather Group), Maxar, and Atmospheric G2 who all do good work in this space. You do this long enough, and you can find reasons to like or dislike any given forecast. Whatever the case, feel free to explore the individual forecasts there, but the consensus (16/7/3) is about as good as it gets right now.

That said, it’s complicated. Tomorrow, in part 2, we’ll talk about El Niño, and what it may mean for the season ahead.