El Niño and the Atlantic hurricane season: why the first hurricane did not arrive until October
The Atlantic's first hurricane of 2026 was Isaias, and it did not arrive until October. Even then the season was far below a normal pace.
The Atlantic hurricane season opens on 1 June. This year, its first hurricane waited until 8 October.
On the National Hurricane Center's 1991–2020 calendar, that first hurricane normally arrives by 11 August, and the first major hurricane by 1 September, in a season that averages 14 named storms, 7 hurricanes and 3 major hurricanes, Category 3 or higher. This year Isaias was not classified as a hurricane until 8 October. By 9 October, Isaias had become the season's first major hurricane, 38 days after the climatological date of 1 September.
Even with Isaias, the season remained unusually quiet. As of the morning of 15 October, NHC's seasonal summary listed 9 named storms, 1 hurricane and 1 major hurricane. It also reported accumulated cyclone energy of 18.0, 83% below the 1991–2020 normal. The season runs until 30 November, so the total tally can still change. What is already clear is how unusually late the season's first hurricane arrived.
This follows Supersized El Niño. The point carried over is that El Niño tends to raise vertical wind shear over the tropical Atlantic, and that lower odds are not zero odds.
Accumulated cyclone energy, or ACE, is NHC's sum of the squares of the maximum sustained surface wind, in knots, every six hours, for named storms while they are at least tropical-storm strength. The Atlantic total on the 15 October summary was 18.0, in units of 10⁴ kt². Of that, Isaias accounts for 7.2.
The NHC summary reports 18.0 as 83% below its 1991–2020 comparison normal. That percentage implies a reference value of approximately 106, although the summary does not explicitly identify the ACE comparison period. NOAA's Climate Prediction Center gives the full-season 1991–2020 average as 122.1, against which 18.0 would be approximately 85% lower. The distinction matters because the season was still in progress. By 15 October the storm-count calendar has normally reached 12 named storms, 6 hurricanes and 2 major hurricanes. This year's totals so far are 3, 5 and 1 below those figures.
In August, NOAA put the chance of a below-normal season, running from 1 June to 30 November, at 75%, with ranges of 7 to 13 named storms, 2 to 6 hurricanes and 0 to 2 major hurricanes, against that average of 14, 7 and 3. The near-normal chance was 20%, and the above-normal chance was 5%. The May outlook was a separate forecast, with a 55% chance of a below-normal season.
El Niño was the reason NOAA gave in August, as covered in the previous article. To see how it can make hurricanes less likely without stopping every one, it helps to look at what a developing storm needs.
A hurricane starts as a cluster of thunderstorms that begins to spin. Warm water from the ocean beneath supplies heat and moisture, fueling the storm. As the moist air rises, it condenses and releases more heat, which pulls more air upward. The winds organise around the falling pressure at the centre, and away from the equator the Earth's rotation gives the inflow its twist, due to the Coriolis force. If the wind changes little with height of air column, the thunderstorms can remain stacked over that spin, allowing the storm to strengthen. Once the surface wind reaches 64 kt, about 74 mph, it is classed as a hurricane.
Light shear and strong shear
Diagram of the general process. Light shear, on the left: the storms stay stacked over the spin. Strong shear, on the right: the storms are blown off the spin. Not to scale. Not Isaias, and not the shear analysed on 9 October.
What breaks the stack is vertical wind shear: wind that changes strongly with height of the air column. The winds aloft carry the thunderstorms away from the circulation underneath, so the heat is released away from the storm's central circulation. A disturbance fails to organise, or a storm that has already formed weakens. Cooler water reduces the oceanic energy supply, while dry air can disrupt the storm's convection, weakening the storm with or without shear.
As discussed in the previous article, El Niño tends to raise that shear over the tropical Atlantic across the season, which shifts the odds against storms and hurricanes. However, that still doesn't tell us where a storm forms or how many there will be in a given season. The water, the moisture and the shear still have to be read where a storm is actually trying to form, when assessing that storm's formation, track and intensity.
On the morning of 7 October, the National Hurricane Center described the water under Isaias as 30–31°C, with shear of 10–15 kt in the SHIPS guidance cited that morning. The storm was still a tropical storm with a broad circulation, over very warm water and with relatively light shear that favoured its development. The forecast that morning took it to a peak of 95 kt within 48 hours. Isaias went beyond that.
By the morning of 9 October it was a major hurricane. An Air Force Reserve aircraft measured 120 kt at 700 mb, about 3 km up, in the eastern eyewall, the ring of strongest wind around the eye. That is not the surface wind NHC uses to rate a storm. From dropsondes, instruments released from the aircraft, NHC estimated the surface wind at 105 kt, about 120 mph, and the central pressure at 959 hPa.
The same discussion put the deep-layer shear at 30–40 kt from the southwest, substantially stronger than the 10–15 kt indicated in the guidance two days earlier. NHC had anticipated stronger shear as a trough, an elongated region of relatively low pressure aloft, approached the northern Gulf.
Direction mattered. Shear across a storm's path pushes the thunderstorms off the centre. NHC noted that this shear was nearly aligned with the storm's motion, so it did less of that. NHC also cited increasing upper-level divergence, air spreading out above the storm. The alignment and increasing upper-level divergence may have helped Isaias maintain its intensity despite the stronger shear, although the storm was not completely immune to its effects. The official forecast at that hour still had 100 kt just offshore the western Florida Panhandle that evening, a step down from the 105 kt estimate.
Later that afternoon, NHC reported that the hurricane had become increasingly asymmetric and that radar observations showed its circulation tilting with height. The stronger shear had not prevented Isaias from becoming a major hurricane, but its effects were becoming visible.
Storm structure and shear
Schematic of the reports in the 7 and 9 October discussions, not a weather map and not to scale. The 10–15 kt shear is the SHIPS guidance cited on the morning of 7 October. The 30–40 kt shear, the flight-level wind, and the 105 kt surface estimate are from the 15:00 UTC discussion on 9 October. The corner marked as not this storm shows shear across a track.
NOAA's hurricane mission deployed small Oshen C-Stars this year, uncrewed surface vehicles that measure wind and pressure at the sea surface. Aircraft and satellites observe different parts of a hurricane, but neither provides a continuous, direct record of surface conditions everywhere. A C-Star can measure those conditions directly as the storm passes nearby, although only at the spot it occupies. That is also what makes getting measurements from a hurricane so difficult: the vehicle has to be in the storm's path before it arrives.
Even in the outer circulation, rather than the eye or eyewall, these measurements offer a rare direct record of conditions as a major hurricane passes. These observations are public NOAA data, and this analysis is independent.
Of the twelve Oshen C-Star datasets publicly available through NOAA's archive at the time of this analysis, PD22 was the only vehicle represented by observations in the Gulf of Mexico. The mission page describes 13 C-Stars. In those twelve datasets, the positions from 4 to 11 October put the other eleven in the western Atlantic or the Caribbean.
Where the C-Stars were
C-Star positions, 4–11 October 2026, from the twelve public NOAA AOML and NOAA PMEL datasets for the 2026 mission. GPS-accepted fixes only. Isaias's track is the same preliminary best track as the Gulf map, drawn here without the wind scale. PD22 is the only Gulf vehicle in that public set. PD20 is not plotted.
Isaias tracked from the southwestern Gulf to the northern Gulf Coast and came ashore on the Florida Panhandle. The colour of each dot is the maximum sustained wind at that six-hourly fix.
Isaias and PD22, 7–10 October
Preliminary NHC best track, 00 UTC 7 October to 12 UTC 10 October 2026, coloured by the six-hourly maximum sustained wind, in knots. PD22 positions are GPS-accepted observations from NOAA's public Oshen C-Star data. The open marker is the centre interpolated along a great circle to 14:44 UTC 9 October.
At 14:44 UTC on 9 October, PD22 was 155 km from the centre. That centre is interpolated between six-hourly fixes, so the distance is not to a reported position. Both fixes had the storm at 105 kt and 959 hPa. The centre moved 174 km in those six hours. According to NHC's analysed wind radii for the northwest quadrant, PD22's closest position was within the estimated tropical-storm-force wind extent but beyond the estimated 50 kt wind extent. Those radii are NHC's storm-wide analysis, not boundaries measured at every point. A quadrant radius is the furthest that wind reached in the analysed field.
Around that pass the gust reached 40.6 kt, and the pressure fell to about 1002 hPa, against 959 hPa at the centre. The vehicle's wind is a mean over the previous two minutes, not the one-minute sustained wind NHC uses to rate a hurricane, and that mean stayed below the gust. In the three hours around the pass the two-minute mean reached 33.7 kt. At the closest observation itself the mean was 25.2 kt, the gust 31.1 kt, and the pressure 1003.4 hPa. At that distance PD22 was in the outer circulation. It did not enter the eye or the eyewall, but it had still managed to record wind and pressure from within the circulation of a major hurricane, something that is remarkably difficult to achieve with a small uncrewed vessel.
Oshen has since confirmed that another C-Star, PD20, was also in the Gulf of Mexico at the time and passed much closer to Isaias, collecting wave, sea-surface temperature and conductivity-temperature measurements. Those observations are yet to become available in NOAA's public archive and once they do they could offer a closer look at the ocean's response to the hurricane.
Wind and pressure at PD22
PD22 wind and pressure, 8–10 October, from NOAA's public Oshen C-Star data. The wind is motion-corrected and averaged over two minutes, not NHC's one-minute wind. The dashed line is the closest approach. The 959 hPa centre is off the scale. Rows with a wind-sensor fault are omitted from the wind panel.
The same afternoon, plotted against distance, shows how narrow the sample is. From noon to 18:00 UTC the vehicle stayed between about 155 and 180 km out. For most of that window it was northwest of the centre, and southwest of it after about 17:00 UTC, so the points before and after the closest approach are not two trips along one line. The region inside 155 km is empty because this record has no measurement there.
One place in the outer circulation
PD22, 12:00–18:00 UTC 9 October 2026. Distance is from the vehicle to the centre interpolated between the six-hourly best-track fixes. Mean wind and pressure use the same two-minute samples as the time series. Circles are before 14:44 UTC. Squares are after.
Sea-surface temperature on PD22 was 29.4°C at the closest observation, and between 28.7 and 29.9°C over 4–11 October. The 30–31°C referred to sea-surface temperatures near Isaias on 7 October, farther south. It is not the water under this vehicle on 9 October.
Among the other eleven vehicles in those public datasets, none came closer than 572 km. Their winds and pressures over those days do not show the core.
At 01:30 UTC on 10 October, an NHC operational update reported landfall near Destin, Florida, with winds of 105 mph and a central pressure of 971 mb. The preceding 00 UTC best-track fix estimated 95 kt and 967 hPa. These estimates refer to different times and should not be treated as competing measurements of the same storm state. The peak used in the track figure is the 105 kt and 959 hPa on 9 October.
Isaias illustrates three challenges in predicting hurricane activity, from the odds of a storm forming to forecasting its strength.
The pattern is not the event. El Niño changes the odds of favourable conditions across a season. NOAA's August outlook gave a 75% chance of a below-normal season, with 0 to 2 major hurricanes, and named El Niño as the reason due to the disruptive shear being more common over the tropical Atlantic. That statement holds for the season as a whole, but it does not say where or when an individual storm will form. Nor can one major hurricane be taken as evidence that the outlook was wrong.
The forecast is not the outcome. On the morning of 7 October, NHC described warm water, light shear and a trough on the way. It expected the shear to rise sharply on the last day before the Gulf Coast, and forecast a peak of 95 kt. Isaias went beyond that. Forecasters can see favourable conditions, and can see that the shear is about to change, without knowing the precise peak. The miss does not, by itself, identify which process, which gap in the observations, or which part of the forecast produced it.
The observations are incomplete. We see only a sliver of the storm. In the public data, PD22 recorded wind and pressure at one place in the outer circulation, 155 km from the interpolated centre at its closest approach. Aircraft, satellites and buoys each sample a different part of the storm, or a different property of it. A forecast begins from those pieces, assembled into an estimate of the state of the whole system. PD22 does not show the eyewall, and it does not by itself show that any forecast improved.
Isaias shows why these distinctions matter. El Niño changed the seasonal odds, but it did not prevent a hurricane from forming. NHC could identify favourable conditions and an approaching trough, but the storm exceeded its initial intensity forecast. In the public data, a vehicle 155 km from the centre could observe only one part of the circulation.
The ocean and atmosphere are constantly influencing one another. Warm water can help fuel a hurricane, but ocean temperatures can also be shaped by the winds and weather patterns above them.
The next piece in the series is the North Atlantic's cold blob, and whether it is a clue to Europe's unusually hot summer.
The Pacific event is Supersized El Niño. NOAA's seasonal statements are the May outlook and the August update. The progress dates are the National Hurricane Center climatology. The storm tally is the season summary. The 7 October environment is discussion 3, and the 9 October aircraft report is discussion 12. The vehicles are described on the 2026 NOAA mission page.
Explore blocking maps · Science and glossary
