Our 2021 hurricane forecast: A BIG season is on the way
By Joe Bastardi
Be prepared! The man who called Hurricane Sandy forecasts a rough 2021 U.S. hurricane season.
I am making this available for our readers so like last year, you can see that we put out the “numbers” in March then follow it up in April with where we believe they are going. This is to show our in-house research on all of this, though we can only show the results rather than get into the nuts and bolts of reasoning here. I will say that the warmer cycle of the Atlantic basin and enhanced MJO reactions in the Pacific ocean seem to be giving us earlier hints, so this is a good test.
The second item, and this is my OPUS, getting a more descriptive scale for hurricanes. I have been working on this for two decades and you can read it below at this link:
- Another big year is on the way.
- There may be 3 predominant tracks this season.
- One each for the eastern and central Atlantic and the other in close.
- Late feedback development will likely be a big forecast challenge again.
- Not as quick a start, but overall impact in worst case could rival last year and back to back challenge 04,05 tandem
Named Storms: 16-22
Major Hurricanes: 3-6
U.S. Landfalling Hurricanes: 3-6
U.S. Major Impact Landfalls: 2-4
U.S. Hurricane conditions: 4-8
For the entire basin, I think that the overall storms will be relatively stronger than normal outside of the Main Development Region (MDR). The naming frenzy, combined with warm water farther to the north, will tend to pull the mean area to the north.
For now, the numbers remain the same as the March issuance. I have added another category, hurricane conditions, as a storm can stay offshore and still blast someone with a hurricane. The definition of landfall is a single point, but storms are not single points. If you are in the eyewall for 12 hours, even if the storm stayed offshore (this has happened quite a lot around Cape Hatteras and can happen with a stalling storm that is paralleling the coast nearby), then you could be having hurricane impacts without a technical landfall. A great example of this, perhaps the best, was Saffir-Simpson Category 4 Helene in 1958, staying offshore but still causing major hurricane impacts on the North Carolina coast. That produced Cape Fear’s strongest gust on record.
While I expect plenty of recurves and northern Atlantic spinups, a more pronounced cluster of tracks may develop. So the Windward Islands will have higher than average activity. Otherwise, the main tropical development areas and islands will be near normal. The only place with below normal conditions will be in the area of Central America that was devastated last year.
The red areas represent a risk of more than 6 times the normal impact. Yellow is 3 times and green is below normal. For example, this means that if the average impact per year in an area is Saffir-Simpson 1.5 (all the years and see what the total rating is at landfall and divide by those years), then that area highlighted in red should get a total of 15. It can occur in multiple storms. Since this example would imply a total of 9, it means it is highly likely a major or two will hit there. Like using a normal snowfall and temperature to compute deviation from average, you must also look at the specific area, as those places to the northeast don’t get hit as much as the Gulf Coast.
I will use last year as a quick and dirty example of what I am talking about. Let’s take the western Gulf of Mexico outlined in the red from last year’s impact forecast:
Talk about an advertisement for our ideas, look at the NOAA warning verification:
This was the activity in the western Gulf of Mexico last year:
Just in that area alone, there were 15 impact points in 2020. Over the years the estimate I came up with as an average (many years there are no hurricanes in here) is 1.5. Last year was 10 times that. So that is the metric I am trying to create to score this.
Now for the sake of argument, why would the season be less active, perhaps much less, than what I am portraying? There are seasons that looked like they should have been much bigger, like 1956, 2007, and 2013, which turned out to have much less activity. Just like this winter looked more like an El Niño at times (despite a moderate La Niña) or the 1951-52 and 2018-19 winters turned out to be opposite of what the ENSO signal gave, you can have a period when there is a large-scale driver that is overwhelming the pattern. The total tropical activity (ACE) of those 3 years above was only around 67! I always make sure I look at what can go wrong. The problem is that though I have identified these before, I have had trouble beforehand in picking them out. So that’s a concern here.
2007 was a case of the ridge being too far to the south over the Southeast. Modeling this year is showing the ridge in the classic banana arc over the top.
These are the 5-highest impact years in the current warm era:
A close-up of the 500 mb forecast:
So right off the bat, the signature is there. The old-timers used to call the positives over the Maritimes the “Newfoundland wheel” because higher than average surface pressures are pulled northward and anything under it would be wheeled into the U.S.
The non-years have the ridge over the Heartland and a trough in the opposite areas:
So that is the opposite of what is being forecasted here. Given the evolution of the spring pattern, I am pretty confident we are not going to see the inactive sample.
Another element that could slow the season down is that we will not be getting the “forecast gift” of the MJO this year. For those of you that read our forecasts last year, you don’t know how many nights I thanked God for that change in the MJO regime we caught so that we could give everyone a heads up about the activity that we were going to see. I did not, nor have I seen, any writeup in previous years jump on the change in the MJO like that. In fact, it was the winter season debacle that actually helped me since when I figured that out and understood the opposite phases for cold were also the strongest hurricane hit correlations, it became clear.
So you cannot ignore the MJO. Phase 4 is the phase that correlates to high northwestern Gulf of Mexico activity.
By and large, the storms hit on the opposite side of the MJO hemisphere that we saw the MJO in during the prior winter.
January-March 2020 MJO:
You can’t get much more opposite.
So to catch a change in that major hemispheric driver meant it was going to hold in some form for several months. Not only do we not have that tool this year, but there is always a danger that pattern is still driving all this now because we are in a long-standing pattern that has been undetected. So the best I can do is try to line up the analog season tropical precipitation with the modeling ideas.
Of late the MJO has been tending to go opposite of last year, but I think that is because there are changes in the Pacific going on that will then level off and allow it to be weaker this year or in Phases 3 & 4. That is not etched in stone.
So you see there are two “negative” influences, but also in this section the countering strong model correlation to it being the opposite at 500 mb to those negative years.
Recently the MJO has been gaining amplitude and a move into those phases in May and June ( 8,1,2,3) could improve on the hints I am seeing from this far out:
European Model ACE forecast
The average of the April European ACE forecasts over the past 5 years has averaged a bit over 70% of normal. Every one of those years they had below normal activity. Its greatest blunder was 2017, where it had about 70% of normal. The average ACE of all these years was 175 or about 160% of the average ACE.
The implications are startling. If they have not changed the way they compute that, then for it to be seeing that large of an ACE forecast when it had below normal activity during two hyper-active years, then it must be seeing something. I suspect they may have made some kind of correction, but it’s very tough to dig into the weeds to find that stuff.
Global precipitation pattern
Here is the forecasted pattern for August-October:
That shows a lot of convection from the Indian Ocean eastward to Indonesia. This would line up with the JMA upward motion forecast for summer along with its precipitation forecast.
It is displaced a couple of months from the heart of the season, though. This lines up with the analog pattern over Indonesia into the eastern Indian Ocean, but also notice the drier western Indian Ocean in the analogs and the JMA showing sinking air over the Gulf of Mexico.
The Euro’s forecast precipitation pattern:
Surface pressure pattern:
The combination of that sinking over the Gulf of Mexico, the precipitation pattern, and the pressure pattern indicates that we are faced with another year of in-close rapid feedback threats. The precipitation pattern above shows a distinct southern wave train and hence the enhanced area versus average over the southeastern Caribbean. Then we see a lot of normal to above in the Atlantic, meaning another congregation of storms that can add to numbers but affect little else but shipping. Then there is the third area near the U.S. coast.
A closeup version of the rainfall pattern:
That is a very close match to the selected analog years of high impact in the current warmer era:
2020 saw a lot of late development. 2017 had Irma that was long-tracked but Harvey blew up in the western Gulf of Mexico. 2008 was a longer tracked year. 2005, while hitting off their peaks, Rita, Katrina, and Wilma all were stronger near 80°W than 60°W. Another year in there is 2011 with Lee and Irene.
Now one more look at the higher than average pressures on the East Coast and ask why it would be higher than average there. Averages might be lower or most storms like even the most active East Coast years will miss the East Coast. So it would only take one big one or a couple of smaller ones. The pattern over the northwestern Atlantic at 500 mb is a problem.
So a lot of recurves are expected but of the storms that are going to deepen, most will be in the last two days as they approach the coast. So it’s a potential double nightmare. It’s not magic, it’s what the pattern dictates.
The May Trough
This theory is not mine. It was put in my head many years ago by a brilliant meteorologist I worked with who will remain nameless. That he saw this in a colder era boggles my mind, since it’s when it’s warmer like this that these tools typically appear more.
If May has a lot of cold air in North America, it indicates higher than average amplitude is likely to occur during the summer season. So there must be a balancing act. It’s why you see so many big storms occurring to the east of where major cold is attacking. For those that have followed WeatherBell over the years, I have pointed out the numerous cases where cold air is showing up in advance of a storm. The reason is sort of intuitive. If there is enhanced surface pressure with cold over the U.S., then there should be enhanced convergence in the tropical waters near the U.S.
The month of May seeds this in two ways. First, it’s indicative of future episodes of amplification during the warm season. Secondly, if there is a negative NAO it slows the easterlies down a bit and allows for more early-season convergence over the warmer waters. While that may not produce any tropical activity then, it does help with the buildup of moisture in the tropical playing field for later in the season.
These are observations, and there is no peer-reviewed research. It’s just something I have used over the years and notice the correlations. So I am offering the reader an explanation. Before dismissing it, I would advise going forth to see what happens over a 30 to 40 year period. As I said, this warmer era is helping me see things that were tougher to see before.
So the kind of May I am expecting would also line up with the hurricane season idea. In fact, the analogs of those four hurricane seasons look like this for May:
So it will be interesting to see if the correlation holds of the cooler East and Plains.
AS if to go to this, the latest Modeling is showing that kind of trough over the NE and lakes. The reasons for this are beyond the scope of this, but it is going to be an interesting test of the warm era correlation ideas as the 4 years I am using looked like this at 500 mb:
The Latest JMA:
While there are some reasons for the season to be less than advertised, there are compelling similarities to some of the other major seasons this century to warrant concern. Our forecast again attempts to quantify the effect, not just list numbers of storms. In the past, I have made the point about dividing the Atlantic hurricane season into two areas, but since that is not going to be done, our initial numbers are in the category of another major season again. Where we are different from the other sources out there is the subject of this missive – the impacts and where they are going to go.
Last year, depending on what scale you use, the impact number on the U.S. Coast was 18-20. This was a record on both the Saffir Simpson Scale and my Power & Impact scale. To put that in perspective, the 2005 season with three major hits and other lesser hits had fewer total impact points on the U.S. coast. Where they hit was obviously a huge key. If two major hurricanes hit near Cameron, Louisiana it is a far cry from two majors hitting New Orleans or the biggest Gulf Coast city, the Houston-Galveston area. There is nothing that says that cannot happen. In fact, you have to wonder why it has not.
It’s that area on the western and central Gulf Coast that is the focus of our highest impact along with the Carolinas. There is also a threat for above average impacts farther to the north up the East Coast and in South Florida. Keep in mind it takes fewer storms and weaker storms to produce above average conditions in New England than it does for South Florida. I have no call for the Caribbean and the Bahamas due to the factors listed above, such as late development, a lot of tracks out to the northeast, and also a cluster to the south, which does put an enhanced area over the southeastern Caribbean.
The end game this year will not be like last year, which actually simply saw the areas where storms developed and hit move southward with time. This year I think Central America will get left alone overall, but that will be a call that I would have no problem updating if I see one coming for them.
With all these things in mind, when it’s all said and done, back to back, this may exceed 2004 and 2005 and become the new benchmark for hurricane hits on the U.S.
Joe Bastardi is a pioneer in extreme weather and long-range forecasting. He is the author of “The Climate Chronicles: Inconvenient Revelations You Won’t Hear From Al Gore — and Others” which you can purchase at the CFACT bookstore. His new book The Weaponization of Weather in the Phony Climate war can be found here. phonyclimatewar.com