Record Baltic sea-level depression exposed unusually strong atmospheric forcing

The Baltic Sea experienced an exceptional basin-wide sea-level depression during winter 2025/2026, with regional sea-surface-height anomalies reaching about 60 centimetres below normal and locally approaching 70 centimetres below normal. The minimum occurred on 7 February 2026 after weeks of persistent high pressure, easterly winds over the southern Baltic and North Sea, and an atmospheric setup favourable to water export through the Danish Straits. A study combining satellite sea level, atmospheric reanalysis, tide-gauge records, modelled transport, ship surveys, Argo floats and gliders found that the event combined some of the strongest antecedent atmospheric forcing in the ERA5 record with the fourth-deepest winter minimum in a 139-winter observational comparison.

The sea-level decline was spatially coherent across the analysed Baltic subregions, including the southern Baltic, Arkona Basin, Gotland Basin, Gulf of Gdańsk, Gulf of Finland and Gulf of Bothnia. Basin-mean sea surface height fell to approximately minus 60 centimetres at the February minimum, while some regions locally dropped below minus 70 centimetres. The common timing and broad pattern indicated a basin-scale barotropic response rather than a local anomaly around one coast or subbasin. After the minimum, sea level rose during late February and March, although it remained below early-winter levels for several weeks. The authors interpreted this sequence as a major reduction in the amount of water stored in the Baltic.

Atmospheric pressure supplied the clearest explanation for the sustained forcing. Mean sea-level pressure over the Baltic repeatedly exceeded 1,020 hectopascals from December 2025 into early February 2026, sometimes covering nearly the entire basin. A revised pressure-persistence index accumulated positive standardized values over a backward-looking 30-day window and reached its event maximum on 20 January. Sensitivity tests using 10-, 15-, 20- and 30-day windows confirmed the importance of sustained high pressure, although the exact date of the index maximum varied with the window length. This was therefore a multi-week anticyclonic episode, not simply the product of one exceptionally strong weather system.

Easterly winds added a coherent outflow-favourable influence. Over the southern Baltic, the strongest lagged association between wind forcing and sea surface height occurred when the wind led the sea-level response by six days, with a correlation of minus 0.78. For the North Sea, the corresponding lag was about one day, and the correlation remained at plus one day under one-, three- and five-day smoothing. The longer-term analysis linked Landsort and Landsort–Norra sea level to the pressure difference between the Baltic and North Sea, but the filtered relationship was moderate: a seven-day lag regression produced a slope of minus 1.33 centimetres per hectopascal, while the pressure gradient explained about 15% of the variance in the sea-level series. The pressure–sea-level relation therefore supported the proposed mechanism without representing a unique adjustment time or explaining the entire event.

Historical comparisons confirmed the event’s rarity. During the 30 days before the 7 February minimum, the easterly wind component over both the North Sea and southern Baltic ranked first among 87 winters. The North Sea component averaged 6.98 metres per second, above its historical 95th percentile of 5.12 metres per second, while the southern Baltic value averaged 5.58 metres per second, above its 95th percentile of 4.23 metres per second. A compound index combining these winds with pressure-gradient persistence ranked second among 78 winters with sufficient data and exceeded its historical 95th percentile. Alternative formulations placed the event consistently second or third. Its tidal sea-level minimum, measured at minus 75.9 centimetres relative to the early-winter reference, ranked fourth among 139 sufficiently complete winters, with the same ordinal result under alternative data-completeness thresholds.

Integrating sea-surface height across the Baltic produced a direct estimate of the associated water-volume loss. The area-integrated anomaly declined from about 12 cubic kilometres above the 1 November–15 December 2025 reference level to approximately minus 290 cubic kilometres on 7 February, a decrease of roughly 302 cubic kilometres. An independent calculation based on net transport across a Kattegat section broadly reproduced this evolution: the transport-derived and sea-level-derived series had a correlation of 0.892 and a root-mean-square difference of 45.3 cubic kilometres. By a spring maximum on 6 April, the sea-level-derived anomaly had improved to approximately minus 61 cubic kilometres, representing recovery of 229 cubic kilometres, or 76% of the preceding decrease. The estimates diverged more clearly after early April.

That recovery cannot be equated directly with an inflow of saline or dense water. The sea-level calculation describes total water-volume change, while the Kattegat calculation estimates exchange through one section; both are also affected by differences in their control domains, atmospheric corrections and unquantified freshwater terms such as river input and precipitation minus evaporation. By 6 April, the hydrographic observations nevertheless showed signs consistent with deep-water renewal in parts of the southern Baltic. April CTD sections revealed plume-like lower-layer water in the Słupsk Furrow with practical salinity locally near 14. In the Bornholm Basin and Słupsk Furrow, relatively saline and oxygenated deep water was observed, whereas the signal was weaker and more diluted in the Gdańsk Basin, where deep salinity was about 12 and oxygen-depleted water persisted.

Glider missions in the eastern Bornholm Basin supplied the clearest time sequence after the minimum. The 10 and 12 practical-salinity contours shoaled from roughly 55–60 metres in early March to about 45–50 metres by mid- to late April along the sampled trajectory. Water with salinity above 14–15 became more extensive below approximately 55–60 metres, and near-bottom salinity locally reached about 17 from mid-April onward. Deep oxygen concentrations were often below 2 milligrams per litre below approximately 60–65 metres during March and early April, but substantially higher values appeared from late April into May. These observations indicate marked, though spatially uneven, oxygenation of the sampled lower layer. Because successive missions sampled different positions, the shifts cannot be interpreted as a basin-wide change at one fixed location.

The findings place the event within the Baltic’s larger outflow–inflow cycle without establishing that the February depression directly caused a Major Baltic Inflow. The pronounced volume loss may have created favourable initial conditions for later exchange, but the subsequent 229-cubic-kilometre recovery combined strait transport and freshwater effects and did not by itself measure saline-water import. The hydrographic evidence was consistent with an inflow-related signal reaching the Bornholm Basin within several weeks, yet the absence of matched immediately pre-event CTD sections farther east prevents a unique reconstruction of the propagation route or attribution of every observed feature to this event. The evidence also showed incomplete deep-water ventilation: relatively oxygenated water appeared in the Bornholm Basin and Słupsk Furrow, while low-oxygen deep water remained in the Gdańsk Basin. The 2026 episode was therefore hydrographically detectable but still evolving, with its final salt import, oxygen impact and formal inflow classification left for further analysis.

Subject of Research: Atmospheric forcing and post-minimum water exchange during the winter 2025/2026 Baltic Sea-level depression

Article Title: Atmospheric forcing of an extreme Baltic Sea-level depression and post-minimum exchange in winter 2025/2026

Article References: Rak, D., Jakacki, J., Muzyka, M., Bulczak, A. I., Przyborska, A., Markowicz, K., Osiński, R., & Kitowska, M. (2026). Atmospheric forcing of an extreme Baltic Sea-level depression and post-minimum exchange in winter 2025/2026. Frontiers in Marine Science, 13, Article 1949814. https://doi.org/10.3389/fmars.2026.1949814

Image Credits: AI Generated

DOI: 10.3389/fmars.2026.1949814

Keywords: Baltic Sea, sea-level depression, atmospheric forcing, water exchange, Danish Straits, Bornholm Basin, deep-water ventilation, ocean circulation, glider observations, Atmospheric, forcing, extreme