The equatorial Pacific warmed a full degree Celsius in roughly a month — weekly Niño-3.4 readings, the benchmark for El Niño strength, hit +1.7°C by mid-June on IRI's measure, up from +0.7°C when the US Climate Prediction Center declared an El Niño advisory on 11 June — and the hurricane forecasters have been cutting their numbers ever since. Columbia's IRI now puts the odds of El Niño persisting through the September-November peak of the Atlantic hurricane season at 100%, with 13 of the 24 models it tracks pointing to a very strong event of +2.0°C or more. CPC gives a 63% chance that the November-January peak ranks among the strongest El Niños since 1950.
For energy markets, that single number is doing two jobs at once. It is suppressing the Atlantic hurricane season that would otherwise threaten Gulf Coast oil, refining and LNG infrastructure through the autumn. And it is loading the dice toward a warm US winter that would sit on gas demand just as the market moves into Q4. The hurricane story and the winter story are the same story, and both run through the tropical Pacific.
This piece walks through what the forecasting agencies actually expect, where the remaining risk is concentrated, and what a suppressed — but not empty — season means for the fourth quarter.
The forecasters have been cutting
NOAA's seasonal outlook, issued 21 May, calls for a below-normal Atlantic season: 8 to 14 named storms, 3 to 6 hurricanes and 1 to 3 major hurricanes, against a 1991-2020 average of 14 named storms, 7 hurricanes and 3 majors. The agency put 55% odds on a below-normal season, 35% on near-normal and just 10% on above-normal, with 70% confidence in the ranges. Accumulated cyclone energy — the season's total wind energy, and the measure most relevant to aggregate disruption risk — is forecast at 45% to 115% of the median.
Colorado State University, whose seasonal forecasts are the market's other reference point, has cut twice. Its April extended-range outlook already sat below climatology at 13 named storms, 6 hurricanes and an ACE of 90 — roughly three-quarters of the long-term averages. The 10 June update took that down to 11 named storms, 5 hurricanes, 2 majors and an ACE of 70. CSU was explicit about the reason: increased confidence in a moderate-to-strong El Niño at the peak of the season, driving well above-normal vertical wind shear across the tropical Atlantic and Caribbean.
The mechanism is the oldest relationship in seasonal hurricane forecasting. El Niño shifts tropical convection toward the central and eastern Pacific, which strengthens the upper-level westerlies over the Atlantic main development region. The resulting wind shear tilts and ventilates developing storms before they can organise. Storms still form in El Niño years; fewer of them survive to become the long-track major hurricanes that do the economic damage.
Two things stop this from being a simple all-clear. First, the Atlantic itself is still warm — NOAA cited slightly warmer-than-normal Atlantic sea surface temperatures and weaker-than-average trade winds as partial offsets to the El Niño signal, and warm water is fuel whenever the shear relaxes. CSU's April read had the western tropical Atlantic warmer than normal with the eastern and central basin slightly cool — a far less explosive configuration than the record-warm basins of 2023 and 2024, but not a cold one. Second, these are June forecasts of an August-to-October phenomenon. NOAA's mid-season update lands in early August, CSU updates in July and again in August, and the ENSO picture is moving quickly enough that a one-degree monthly change in the Pacific is currently the norm rather than the exception.
Where the storms would go
For infrastructure risk the count matters less than the track, and here CSU publishes the numbers that matter. Its June update puts the probability of at least one major hurricane landfall on the continental US at 24%, against a 1880-2020 average of 43%. For the Gulf Coast specifically — the Florida Panhandle around to Brownsville, the stretch that contains essentially all US LNG export capacity and the largest refining complex in the western hemisphere — the June figure is 14%, versus a 27% long-term average. The East Coast plus peninsula Florida sits at 11% against 21%. The odds of a major hurricane tracking through the Caribbean, often the staging ground for the strongest Gulf systems, are 26% against a 47% average.
Read those numbers carefully: they are roughly half of climatology, not a tenth of it. A 14% chance of a major hurricane landfall on the Gulf energy corridor is the kind of probability that headlines call low and risk managers call material. The 2026 season's tail is thinner than usual. It has not been removed.
The Pacific side of the ledger
El Niño does not suppress storms everywhere — it redistributes them. In the western North Pacific, the world's most active basin, reinsurance broker Guy Carpenter's May outlook calls for a near-normal number of typhoon formations but an above-normal landfall risk for Japan and Korea, with below-to-near-normal risk for the Philippines. That is the classic El Niño signature: storms forming further east and recurving north, away from the South China Sea and toward the Japan-Korea corridor. El Niño years have also tended to produce higher overall western Pacific activity and ACE even when formation counts stay near normal.
For energy markets the transmission is on the demand side. Japan and Korea are two of the world's three largest LNG importers, and typhoon landfalls there disrupt ports, regasification schedules and power systems in the middle of the shoulder season when Q4 procurement is being set. A quiet Gulf of Mexico and a busy Japan-Korea track would be an unusual but entirely consistent 2026 outcome — and one that would matter more to JKM than to Henry Hub.
One caveat belongs here: the Japan/Korea call is a qualitative output of Guy Carpenter's model, it applies to landfall risk rather than basin-wide counts, and equivalent verified 2026 outlooks for the East Pacific and Indian Ocean were not available for this review — though climatologically, El Niño enhances the East Pacific basin, which raises the separate question of Mexico's west-coast exposure.
The Gulf's risk has moved onshore
The most important structural fact in the hurricane-energy relationship is how asymmetric the Gulf of Mexico's exposure has become. On EIA figures, the federal offshore Gulf supplies around 13% of US crude oil production in 2025-26, down only modestly from 14% in 2023. Offshore gas is a different world: the Gulf produced 1.79 Bcf/d of marketed natural gas in 2024, falling to a forecast 1.64 Bcf/d in 2026 — roughly 1% of US output. When Katrina and Rita tore through the basin in 2005, that share was 17%.
The consequence is that a hurricane can no longer create a US gas supply shock by shutting in offshore platforms. What it can do is hit the demand side and the export machine, because the gas story moved onshore just as the LNG buildout concentrated it on one stretch of coast. Sabine Pass and Calcasieu Pass sit on the Louisiana-Texas border, Freeport and Corpus Christi on the Texas coast, Plaquemines below New Orleans — five terminals on precisely the coastline CSU's 14% Gulf landfall probability describes. A major hurricane into that corridor tends to be bearish Henry Hub in the first instance, because an offline export terminal strands feedgas in the domestic market faster than offshore shut-ins remove supply. The bullish leg shows up in Europe and Asia: TTF and JKM price the lost cargoes. Crude keeps the old-fashioned exposure — around one barrel in eight of US production is still produced offshore in the storm path, alongside the Gulf Coast refining complex and the transmission grid that keeps it running.
That reversal of sign — Gulf hurricanes as a domestic gas demand event and an international supply event — is still underweighted in casual commentary, and 2026's export capacity is the largest it has ever been. The same storm that would have spiked Henry Hub in 2005 would more likely sink it today while lifting the European curve.
What it means for Q4
Put the pieces together and the probability-weighted hurricane premium for Q4 energy prices is below normal this year. Every agency in the verified set — NOAA, CSU, CPC, IRI — is pointing the same direction, and the driver is a Pacific event that strengthens through the exact window in which Atlantic risk normally peaks. The forecast revisions have all been one way.
The strong-El Niño analogue seasons, 1997-98 and 2015-16, delivered quiet Atlantics followed by warm US winters and weak winter gas prices. That is the pattern the current forecast distribution points toward, and it is why the bigger Q4 story is probably not the hurricane season at all but the winter that follows it: a very strong El Niño tilts the odds toward mild US heating demand at the same time as it thins out autumn supply risk. For gas, the two effects stack. The caveat on the analogues is worth stating plainly — the US now exports gas on a scale those winters never saw, which changes how much a warm winter can loosen the domestic balance before the export arb takes the slack.
Against that, the tail. NOAA's own below-normal forecast still contains one to three major hurricanes, and it takes exactly one, in the wrong place, on a coastline that now hosts the largest LNG export complex in the world. Terminal strikes are binary events; a below-normal season average says nothing about them once a storm is inside the Gulf. The suppressed-season call is a statement about frequency, not about consequence.
What to watch from here
The next hard data points arrive in sequence. CSU updates in July and again in early August; NOAA's mid-season outlook lands around the same time. Both will show whether the June cuts went far enough — or whether the warm Atlantic argues for a partial reversal.
The weekly Niño-3.4 readings are the single most informative series between now and then. The event is currently strengthening at a pace that has surprised the agencies; whether it stalls in the strong range or runs to very strong (the +2.0°C threshold that 13 of 24 IRI-tracked models now reach) determines both the shear over the Atlantic in September and the shape of US winter demand in December.
Within the season itself, the things that matter are the ones the seasonal forecasts cannot see: whether observed wind shear over the main development region and Caribbean verifies as high as forecast; any period in which the pattern relaxes for two to three weeks — the sub-seasonal windows in which suppressed seasons produce their exceptions; and, on the Pacific side, whether the Japan-Korea recurvature signature shows up in the early typhoon tracks.
The climatological peak of the Atlantic season runs from late August through the first week of October. In a suppressed year that window narrows but does not close. The market read for now: fade the season, respect the tail, and watch the Pacific — it is setting the price of both risks.