This year’s Atlantic hurricane season is proving to be one of the least well predicted of the 21st Century to date. Not in terms of how active tropical cyclones are behaving – the NHC are going a fantastic job there as usual – but how active the season is overall.
What follows here is a deep dive into those predictions and the surprising reality through late September, rounding off with a brief take on the prospects for October 2024.
The left-hand table below shows the seasonal predictions for 2024 released by a wide variety of institutions.
They’re unanimous for a more active than usual season when judged by the numbers of named storms (TS+), hurricanes (C1+), and major hurricanes (C3+).
Among the most dramatic predictions came from MU, SMN, and NOAA, each indicating a high chance of at least 20 named storms – twice the 1981-2020 average - and 10 hurricanes – 3 more than average.
In April, UPenn predicted an astonishing 33 named storms!
At the other end of the spectrum, MetSwift’s December-issued prediction was for ‘only’ 17 named storms and 8 hurricanes.
The actual behaviour of this season through September 23rd has been different to say the least. To this point, 1981-2020 typically saw 9-10 named storms, 4-5 hurricanes, and 2 major hurricanes. This year has seen 7 named storms, 4 hurricanes, and 1 major.
The right-hand table above shows approximately what each institution’s predictions look like when adjusted for the season up to 23rd September.
Unsurprisingly, all the predictions are above the observed, with MetSwift closest for named storms and hurricanes, and NSCU closest for major hurricanes.
The by-month breakdown below illustrates that the season sprung into an action with a busy June, only for July to be devoid of activity, August unremarkable, and September so far on the quiet side.
The season is far from done and may yet finish above average (I’ll get to that later), but it looks likely that most of the predictions will prove to have been at least a little too high.
To put it shortly, the unanimous prediction of an above normal season was primarily based on the expected coexistence of a very warm tropical-subtropical North Atlantic (and in many cases Caribbean Sea) with an event in the tropical Pacific known as La Niña.
This combination has historically encouraged busier than usual Atlantic hurricane seasons, including some of the most active on record.
If one accounts for long-term warming skewing traditional measures of the El Niño Southern Oscillation, then this combination of La Niña and Atlantic-Caribbean warmth has very much materialised. Yet the busy hurricane season hasn’t, at least so far.
I’ve seen many proposed explanations doing the rounds, among which the three most compelling are a climate trend toward a more stable atmosphere over the tropics, an exceptionally north-shifted path of tropical waves over Africa, and an extremely unusual circulation pattern over the North Atlantic.
It’s well established that temperatures in the troposphere – the layer of the atmosphere closest to the surface – have warmed substantially over the past century. Also, that this warming leads to higher typical tropospheric moisture content, as increased evapotranspiration supplies additional moisture to the air, which itself has a greater holding capacity.
Importantly for tropical cyclones, the more moisture air contains, the more slowly its temperature drops with height. So, as can be seen in the plot below, when the whole troposphere is warmed, the change is larger in the upper levels (200 mb air pressure level) compared to lower down (1000 mb).
During 1980-2009, the troposphere at 200 mb was on average 79.4°C cooler than at 1000 mb. 2010-2024, that difference has reduced to 79.0°C. This may not sound like much at face value, but the difference has only varied within a range of 1.5°C across 1980-2024, meaning the recent reduction is over quarter of the historical range.
So, temperatures now tend to drop more slowly with height, which means the atmosphere is less unstable. In short, as warmed air parcels rise and cool, they tend to lose buoyancy at a lower altitude. To reach the same height as before now requires a higher starting temperature, greater moisture content, or both.
Unfortunately for tropical cyclones, moist air parcels rising very high in the troposphere, forming towering clouds (deep convection), is fundamental to their development. Now, unless short-term variability leads to a cooler troposphere (as seen in Jun-Aug 2023), a greater surface heating plus convergence of moisture is required for a tropical cyclone to get going.
Not only that, but the more stable atmosphere makes it easier for dry, dusty air, riding the trade winds from the Saharan desert, to descend into tropical disturbances or cyclones and deteriorate their structure. It prevents at least a few would-be tropical cyclones in most years, but the hit rate has been unusually high during Jul-Sep 2024.
So, that’s one of the major factors behind the below-par activity of late. Next, the waves of Africa - not oceanic waves, but atmospheric ones.
Every year, between July and the first half of autumn, intense thunderstorm complexes march westward across the African tropics. They represent mesoscale organisation to thunderstorm activity associated with a large-scale feature known as the inter-tropical convergence zone.
This year, it has shifted exceptionally far north during Jul-Aug and still hasn’t returned to its usual position as of late Sep.
As detailed in my previous blog entry, this has led to an unusually soggy Sahel (sub-Saharan Africa), where wide swathes have seen a dramatic increase in active vegetation (mainly grasses and similar plants).
Now here’s the kicker: These waves typically continue across the tropical Atlantic and are the most common ‘seed’ for tropical cyclones in Jul-Sep, even as far west as the Caribbean.
However, to develop into a tropical cyclone, the waves must not be taken apart by horizonal wind shear or dry air intrusion. In 2024, the waves have travelled unusually close to the Sahara, so have tended to ingest more dry air, leading to their collapse.
The combination of this with the more stable troposphere is already enough to put a dampener on Atlantic hurricane seasons, but wait, there’s more!
The summer of 2024 was extremely unusual across the northern half of the North Atlantic. From southern Greenland to northern Scotland, a measure known as geopotential height was very far below normal. This corresponds to a much greater than usual frequency of low pressure systems within that area.
Meanwhile, the opposite was true within a region spanning Bermuda and the Azores. As shown below, the difference between geopotential heights between the northern and southern regions was the most negative since at least 1979.
Importantly relative to the Atlantic hurricane season, this pattern makes for a very strong and north-shifted ‘subtropical high’, centred around 45N.
Highs in the northern hemisphere have a clockwise circulation, which means the Atlantic subtropical high has a boosting effect on the trade winds heading west-southwest from Africa.
So, with a stronger high, you get stronger trade winds, which transport a larger amount of dry, dusty air from the Sahara to the tropical Atlantic.
That bring us to three suppressive influences on Atlantic hurricane season activity. Which begs the question, why have all the forecasts for this year called for a busier than average season?
Throughout 2023 and 2024 so far, sea surface temperatures in the tropical Atlantic have been near or at record-high levels. Seasonal models correctly predicted the continuation of impressive warmth during Jul-Aug (see below).
Would this be enough to outweigh the suppressing effect of upper troposphere warming on tropical cyclone genesis? With no real historical precedent, this was essentially unknown, giving no confident basis for overruling the well-established correlation of higher tropical Atlantic SSTs with more active hurricane seasons.
Meanwhile, the northward shift of the wave train over Africa was also flagged up by seasonal models, but only at 40-60% probability for much of the Sahel (see below). The stronger signal was for those waves to be anomalously strong, which tends to raise the likelihood of them serving as ‘seeds’ for tropical cyclone genesis. So, this too steered forecasts toward a busy Atlantic hurricane season.
Similarly, the seasonal models leaned the right way regarding the North Atlantic pattern, most so across Greenland, but it was a weak signal. Jan 2024 saw a ‘sudden stratospheric warming’ event, which when occurring at that time of year, has a historical correlation with low sea-level pressure around Greenland the following summer. Clearly this steered the models in the right direction, but not with enough emphasis.
Through all of this runs a common thread: High uncertainty when dealing with extreme weather patterns that have little to no historical precedent. Important lessons will be learned from how this season has played out so far.
What of its remainder, then – will it ever kick back into high gear?
During the middle stages of autumn, the upper troposphere cools faster than the oceans, due to water having higher specific heat capacity than air. Meanwhile, the wave train over Africa heads southward and weakens, with a larger proportion of tropical cyclones instead arising from initially non-tropical systems acquiring tropical characteristics.
So, we can confidently expect two of the three suppressing factors to subside, but what about the North Atlantic pattern? That’s a lot more uncertain, with seasonal model predictions for October 2024 providing no clear signal (see below).
However, the absence of any signal for lower than usual sea-level pressure around Greenland may count for something. Already, September has seen an increased number of days with high pressure near or over Greenland, instead of incessant lows.
What’s more, a phenomenon known as the Madden-Julian Oscillation, which can help spin-up tropical cyclones, is forecast to be located favourably for the Atlantic basin during the next 2-4 weeks.
Considering all of this, its hard not to expect a substantial increase in activity across the basin, with above normal numbers of tropical storms and hurricanes during October 2024.
We’re already seeing signs of that at the time of writing (26th September), as powerful hurricane Helene heads toward the south-eastern USA, while tropical storm Isaac roams the high seas to the northeast of Bermuda, and another tropical cyclone is likely to develop to the east-northeast of the Caribbean within the next few days.
James Peacock MSc
Head Meteorologist at MetSwift
Featured photo by Emily Morter on Unsplash
