The North Atlantic's cold blob - a clue to Europe's unusually hot summer?
While European seas hit summer records in 2026, a subpolar North Atlantic patch stayed cooler than average. Whether that cold water helps explain the heat is still open.
While much of Europe's surrounding seas baked this summer, one stubborn patch farther west in the subpolar North Atlantic remained unusually cold. The anomaly had been visible since at least December 2025, well before summer began.
Could that cold patch help explain Europe's unusually hot summer?
Both areas are displayed on the map below, along with their sea-surface temperature anomalies. The warm European seas help explain why coastal nights stayed hot once high pressure was present overhead. The cool patch farther west asks a separate question: whether a pattern that was already in the ocean before the summer season changed the odds of the circulation that kept returning.
It follows Europe's hot summer, where the cold patch first appeared as an open question on persistence, and the hurricane season piece, which asked how the ocean and the atmosphere change each other.
North Atlantic and European seas, summer 2026
June–August 2026 sea-surface temperature, minus the 1991–2020 June–August mean. European seas are warmer than average; the dashed outline marks the approximate cool subpolar patch sometimes called the North Atlantic cold blob — not a contour of the data. Shared ±2.5°C colour scale with Figures 3 and 4; values outside saturate. Land and sea-ice cells are masked. Copernicus has not attributed the repeated blocking to the cool patch. 1991–2020: WeatherBench2 ERA5. 2026: ARCO ERA5/ERA5T. Generated using Copernicus Climate Change Service information 2026.
Over the European ocean domain, 56% of the area was much warmer than average in June–August 2026, and 22% had its warmest summer sea surface in the ranking Copernicus uses, which runs from 1979. Much warmer than average means warmer than 90% of the summers in 1991–2020. The records were concentrated on the southern North Sea, the English Channel, the Irish Sea, the Bay of Biscay, the waters west of Portugal and the western Mediterranean.
The domain is the review's, not the whole North Atlantic: 25°W–40°E, 34°N–72°N, oceans only. The cool patch in the subpolar gyre lies mostly west of 25°W, so the 56% does not describe it, and it does not describe the basin.
The daily averages are absolute temperatures, not anomalies. The Bay of Biscay reached 22.6°C on 13 August. The western Mediterranean reached 29.4°C on 15 August. Large parts of that basin also exceeded 30°C in July, August and September. Copernicus calls the Biscay and western Mediterranean values record highs for those seas in August.
The cool side of the map has no matching number. The summer review says a large area farther west, broadly the subpolar gyre, including the region sometimes called the North Atlantic cold blob, remained cooler than average, and that those anomalies had persisted since at least December 2025. Cooler than the 1991–2020 average is the claim. A single temperature for the patch is not in the review.
Some of the same atmospheric conditions that heated the land also helped warm the surrounding seas. Under persistent high pressure, clearer skies allow more solar energy to reach the surface, while lighter winds reduce mixing with cooler water below. Once that surface is warm, it feeds heat and moisture back into the air along the coast: higher humidity, heavier heat stress, and nights that stay warm. The record warmth coincided with strong-to-extreme marine heatwaves in several European seas, and with extreme conditions in parts of the Bay of Biscay in July.
Tropical nights, when the temperature does not drop below 20°C, reached as far north as northern Germany, the Netherlands and southern Britain. Parts of northeastern Spain and Italy had 30–45 more of the tropical nights than their 1991–2020 summer average. Copernicus links that excess to the record sea temperatures in the Mediterranean and along the coastal northeastern Atlantic.
That local loop explains why the nights did not bring cooling relief, but it does not explain why the persistent high-pressure systems kept coming back. That question is the one the summer review left open.
Farther west, the subpolar North Atlantic tells a different story. In that basin, winter storms do more than cool the surface. Cold air and strong winds draw heat from the sea surface into the atmosphere: this is ocean heat loss. At the same time, strong winds and surface cooling help mix the upper water column, spreading the cooling down through hundreds of metres rather than leaving it concentrated at the surface.
By spring and summer, much of the sun's warming remains concentrated in a shallow surface layer, while a thermocline helps isolate the colder water stored below. When the mixed layer deepens again in autumn and winter, some of that cold water can return towards the surface. Oceanographers call this re-emergence. That seasonal sequence is drawn below.
On its own, re-emergence would not explain a cool surface through June–August 2026, because it happens when the mixed layer deepens in autumn and winter. It is also not an established heat-budget result for last winter's patch.
Subpolar ocean mixing
Illustrative schematic, not to scale. Winter storms take heat from the ocean and mix the cooling down; spring and summer leave that cold water below the thermocline; later deepening can bring it back up. Not a diagnosis of the 2025–26 cool patch, and not by itself an explanation of a cold June–August surface.
The shelf seas around Europe are shallow by comparison. A few weeks of sun and light winds are enough to warm them. Two anomalies on one map can have different ages.
A useful historical comparison is 2015, in Duchez et al.. They used NCEP–NCAR reanalysis from 1948 to 2015, with the 1981–2010 seasonal cycle removed. That is a different dataset and a different baseline from the 1991–2020 ERA5 anomalies in the figures here.
They attributed the cold North Atlantic anomaly to extreme ocean heat loss over the preceding two winters and the re-emergence of colder subsurface water. A cold surface feature was already visible in November 2014: itself a re-emergence of a cold anomaly from the winter of 2013–14, with heat loss over the following winter adding to it. In June 2015, temperatures were up to 2°C colder than normal over much of the subpolar gyre — a system of circulating surface currents — and in places the coldest June of their record. Their Atlantic box is 45°N–60°N, 40°W–15°W. Wind-driven upwelling and a reduced northward ocean heat transport were not the primary drivers in their account.
June 2015 sea-surface temperature anomaly
June 2015 sea-surface temperature anomaly relative to the 1991–2020 June mean. Dashed ellipse: 45–60°N, 40–15°W — the Duchez et al. cold-anomaly box, not a contour of the data and smaller than the full subpolar gyre. Same domain and ±2.5°C colour scale as the summer map. Duchez et al. used a 1981–2010 baseline. WeatherBench2 ERA5. Generated using Copernicus Climate Change Service information 2026.
Duchez and colleagues also examined the five hottest Central European summers since 1980 — summers in which their Central European temperature, over 45°N–53°N and 1°E–35°E, exceeded one standard deviation: 1992, 1994, 2003, 2012 and 2015. When they averaged the North Atlantic sea-surface temperatures from the May preceding each summer, a pattern of unusually cold water emerged in the subpolar North Atlantic. They proposed that the resulting north–south temperature contrast might influence the jet stream and favour high pressure over Europe, but emphasised that this mechanism still needed to be tested with atmospheric models. For 2015 itself, the north–south contrast in sea temperature strengthened from mid-May, and their measure of the jet shifted south and stayed there from late June. The mechanism they suggested was a Rossby wave train set off by that contrast. They did not call that a cause.
The jet stream is partly maintained by large-scale temperature contrasts in the atmosphere. Ocean temperatures can influence those contrasts through heat exchange with the air above. Changes in those gradients may affect storm tracks and atmospheric waves, potentially altering the position or behaviour of the jet stream. But whether a particular ocean anomaly favours high pressure over Europe depends on the wider atmospheric circulation.
What the recent bulletins do show is that the cool water of 2025–26 was not made by the June sun.
In the 2025 ocean review, the central and western North Atlantic were the exception to a record-warm year for the European ocean region: annual sea-surface temperatures there were near or below the 1991–2020 average, and cooler-than-average water spread across that part of the basin in the second half of the year. The European-region annual value, 10.94°C and 0.65°C above average, is a different domain and a different season from the summer map. It does not belong in the same sentence as the 56%.
The December 2025 bulletin is more specific about the mixture. A large region of the subtropical and northeast North Atlantic, including the Norwegian Sea, had the warmest December sea-surface temperatures on record for the time of year. Parts of the North Atlantic were cooler than average. The bulletin does not call those parts a cold blob, and it does not give them an area mean. The summer review is the source that looks back and says the cool anomalies then persisted.
By February 2026 the western and central North Atlantic were still below average, while a large region of the subtropical and northern North Atlantic was at record warmth for the month. Copernicus reported that the relatively pronounced SST gradient was a contributing factor to the active storm track that winter. The highlight on the same page says the gradient likely favoured storms that reached Europe. Either wording is about winter storms. A summer of blocking highs is a different outcome, even if both can be sensitive to the north–south contrast in the ocean.
February 2026 sea-surface temperature anomaly and storm-track activity proxy
February 2026 sea-surface temperature anomaly relative to the 1991–2020 February mean. Contours: storm-track activity proxy — standard deviation of 6-hourly 500 hPa height about the February 2026 mean (gpm). Cooler water in the western and central North Atlantic beside warmth farther south and north. Same domain and ±2.5°C colour scale as the summer map. ARCO ERA5/ERA5T. Generated using Copernicus Climate Change Service information 2026.
Winter storm tracks and summer blocking involve different atmospheric dynamics. In winter, strong temperature contrasts can support the development of storms along the Atlantic storm track. In summer, changes in atmospheric temperature gradients, the jet stream and large-scale waves may influence where persistent high-pressure systems develop.
What about summer 2026? Copernicus documented the persistent cool anomaly in the subpolar North Atlantic, visible since at least December 2025. It says this area broadly corresponds to the subpolar gyre and includes the region sometimes called the North Atlantic "cold blob". Copernicus also says North Atlantic temperature patterns may have influenced the conditions associated with European blocking, and that the causes of the blocking's unusual persistence remain under investigation, including possible influences from those patterns.
I have not found a published heat-budget analysis of this summer's cool patch, or a modelling study that attributes the blocking to it. Until one exists, both the patch's cause and its atmospheric influence remain open. Re-emergence is the 2015 explanation, available as an analogy for how a cool surface can inherit winter conditions, and not yet a result for 25/26 winter.
The research that makes the "may have" concrete is about other summers, and it has a direction.
Ossó, Sutton, Shaffrey and Dong, in 2020, followed a spring pattern in North Atlantic sea temperature through the following summer, in ERA-Interim for 1979–2017 with the long-term trend removed. Their index is a northwest box, 42°N–52°N and 52°W–40°W, minus a box to the southeast, 35°N–42°N and 35°W–20°W, in March and April. When the northwest is warm relative to the southeast, the summer jet sits further north. The cold-north pattern is the other sign, and it is the one that goes with a jet shifted south. That is the same sense as the 2015 hypothesis.
The relationship also works in reverse. Ossó and colleagues identified a reinforcing interaction between North Atlantic sea-surface temperatures and summer atmospheric circulation: ocean temperature patterns can influence the jet stream, while the resulting changes in winds, surface heat exchange and mixing alter sea temperatures in return. By midsummer, the observed cold SST anomaly may therefore reflect both conditions inherited from earlier months and the atmospheric circulation that developed during summer. The map alone cannot separate the two. That feedback is their North Atlantic pattern, centred on the open ocean rather than on the Bay of Biscay. The warm shelf seas have a separate account, the one in the Copernicus review: the highs heated them. The cold subpolar water is the part of this summer's map that was already in place in winter.
A later composite study is sometimes quoted as if cold water preceded heat almost every time. Krüger, Kjellsson, Pilch Kedzierski and Claus, using ERA5 for 1979–2019 with the trend removed, defined European heat events over a central box, 0°–20°E and 45°N–52.5°N, and the Atlantic average over 15°W–40°W and 45°N–60°N. A falling North Atlantic anomaly preceded 15 of their 18 heat events, and cold conditions were present during 14 of the 18. The trend had been removed, and the series stops in 2019. It is a lagged statistical relationship, of the kind Duchez asked a model to test. It is an account of those 18 events, and 2026 is outside it.
The North Atlantic "cold blob" is also the name often given to a longer-term cooling pattern south of Greenland and Iceland, which overlaps with the subpolar gyre. Researchers have investigated whether changes in the Atlantic Meridional Overturning Circulation (AMOC) contribute to this longer-term cooling. AMOC transports warm water northward near the surface and colder water southward at depth.
North Atlantic Circulation and AMOC
Illustrative schematic, not to scale. (a) Subpolar gyre surface circulation; the dashed region marks longer-term cooling often nicknamed a cold blob, distinct from the 2026 seasonal anomaly. (b) AMOC: warm surface water north, heat loss and sinking at high latitudes, cold deep return south. Causes of longer-term North Atlantic cooling, including any AMOC role, remain debated. Not a diagnosis of the summer 2026 cool patch.
The 2025 ocean review records longer-term cooling in that region beside the faster warming of the European seas. Ocean heat transport, the overturning circulation, and heat lost at the surface have all been proposed, and the literature has not converged. A trend measured over decades and a cool patch that Copernicus can date to last winter can occupy similar water. Each needs its own evidence. The summer 2026 anomaly alone cannot establish whether AMOC played a role. This review is about the patch on the summer map.
The seas around Europe were exceptionally warm in summer 2026 because the blocking highs left them under clear skies and light winds, and that warmth then limited nighttime cooling along the coast.
The subpolar North Atlantic was cooler than average, and had been since at least December 2025, so it was in place early enough to matter for the circulation that kept returning. The studied pathway runs in a particular direction: a colder north Atlantic, relative to the ocean further south, has been associated with a southward shift of the summer jet. Copernicus has left the 2026 case open.
The question is not simply whether the cold Atlantic appeared before Europe's hot summer. It is whether the atmosphere would have behaved differently without it. To answer that, we would need to compare model experiments with and without the observed ocean temperature anomaly, and see whether European blocking becomes more or less likely. Until then, the cold patch is an interesting observation, not an explanation.
The next review takes the atmosphere on its own: When High Pressure Won't Move - The Physics of Atmospheric Blocking. If the ocean is only part of the story, what does the atmosphere itself have to do before that stuck weather counts as a block?
Figures 1, 3 and 4 use ERA5 reanalysis sea-surface temperature. Figure 1 is the June–August 2026 mean of the 00, 06, 12 and 18 UTC fields, minus the 1991–2020 mean of the same hours; 2026 from ARCO ERA5/ERA5T. The dashed outline on Figure 1 marks the approximate cool subpolar patch sometimes called the North Atlantic cold blob; it is not a contour of the data. Figure 3 is June 2015 minus the 1991–2020 June mean (WeatherBench2). The dashed ellipse marks 45–60°N, 40–15°W, the region Duchez et al. used for the cold anomaly — not a contour of the data, and smaller than the full subpolar gyre. Figure 4 is February 2026 minus the 1991–2020 February mean (ARCO ERA5/ERA5T), with a storm-track activity proxy: the standard deviation of 6-hourly 500 hPa geopotential height (÷g → gpm) about the February 2026 time mean. Contours show elevated values of that proxy; there is no schematic storm-track arrow. This map does not show whether the track was more active than normal — Copernicus is the source for the “active storm track” claim.
Domain for all three: 20°N–70°N, 70°W–40°E. Shared colour scale ±2.5°C (values outside saturate). Land and sea-ice cells are masked. ERA5T is the preliminary near-real-time extension of ERA5, so 2026 values may change slightly in the final release. All three figures use a 1991–2020 baseline. Duchez et al. used NCEP/NCAR reanalysis with a 1981–2010 baseline, so their 2°C figure is not directly comparable with the colour scale of Figure 3.
Contains modified Copernicus Climate Change Service information 2026. Neither the European Commission nor ECMWF is responsible for any use that may be made of the information it contains.
The season those highs produced is Europe's extraordinarily hot summer of 2026. The piece that pointed here is El Niño and the Atlantic hurricane season.
The summer numbers and the open question on persistence are the Copernicus summer review. The 2015 cold anomaly is Duchez et al. 2016. The spring-to-summer jet relationship is Ossó et al. 2020. The composite count is Krüger et al. 2023. The winter gradient is the February 2026 bulletin, and the longer cooling is the 2025 ocean section.
Explore blocking maps · Science and glossary
