For nearly three decades, research vessels from Russia, Germany, and the United States crisscrossed one of the Arctic Ocean’s most intensively studied seas, deploying temperature and salinity sensors from surface to seafloor, publishing peer-reviewed papers, and building what scientists believed was a comprehensive picture of how water moved through the region. The entire time, a coherent, kilometer-scale ocean current carrying warm Atlantic water along the Kara Sea’s northern margin was flowing right through their data — and no one recognized it as a current.

That current now has a name. Published Tuesday in Frontiers in Marine Science, a peer-reviewed study by Alexander Osadchiev and colleagues at the Shirshov Institute of Oceanology of the Russian Academy of Sciences formally names the Keldysh Current — a surface flow in the Kara Sea that carries Atlantic-origin water from the northern shore of Novaya Zemlya northward through the St. Anna Trough and eastward along the continental slope toward the Laptev Sea. It was named after the Akademik Mstislav Keldysh, the Russian research vessel that has conducted oceanographic surveys in the Barents and Kara seas since 1989.

The discovery is not merely a naming exercise. The Keldysh Current transports roughly 0.4 Sverdrups (Sv) of warm, salty Atlantic water — a Sverdrup being one million cubic meters per second (approximately 264 million US gallons per second), and 0.4 Sv being roughly twice the Amazon River’s peak discharge. It is an estimated 50–100 km (31–62 miles) wide, 50–100 m (164–328 ft) deep, and more than 1,000 km (621 miles) long. The reason it went undescribed despite being measured in 25 shipboard surveys conducted between 1995 and 2024 by vessels from multiple countries — including the German icebreaker Polarstern — is a story about the structural limits of how oceanography detects currents, and what those limits mean for projecting Arctic warming.

What makes the Keldysh Current’s decades-long invisibility significant extends beyond a cataloging gap: climate models projecting how Atlantic heat reaches the Eastern Arctic at the surface have been unable to account for a transport mechanism they did not know existed, and the current’s presence as a surface-layer pathway for Atlantic heat may mean those models have systematically underestimated warming in the Kara Sea, a possibility highlighted in the context of ongoing Arctic Atlantification trends documented by NOAA’s 2025 Arctic Report Card.

How Standard Tools Missed a Current Twice the Size of the Amazon

The core reason the Keldysh Current went unrecognized is a fundamental property of the instrument most responsible for mapping global ocean circulation: satellite altimetry.

Satellite altimeters measure sea surface height by bouncing radar signals off the ocean surface. Oceanographers then use a principle called geostrophic balance — the equilibrium between the ocean’s pressure gradient and the Coriolis force from Earth’s rotation — to infer currents from those height measurements. The world’s major ocean currents, including the Gulf Stream, the Kuroshio, and the Antarctic Circumpolar Current, are approximately in geostrophic balance, which is why satellite altimetry has been so powerful for mapping global circulation since the early 1990s.

A geostrophic current creates a measurable sea surface height anomaly and a corresponding pressure signature in the water column’s density structure. Both of those signatures are what standard oceanographic tools look for. A non-geostrophic current — one that is not in pressure-gradient balance, but is instead driven by entrainment from a deeper flow — creates neither. It is, in a precise technical sense, invisible to altimetry and to the conventional density-based methods ships have long used to reconstruct circulation.

The Keldysh Current is non-geostrophic. It is not driven by its own pressure field. Instead, it is pulled along the pathway of a much more vigorous subsurface current — the Barents Sea branch water (BSBW) — by dynamic entrainment, the process by which a powerful deep flow drags overlying water masses along its isobar-hugging trajectory. Because the Keldysh Current leaves no independent pressure signature, it could not be inferred from sea surface height measurements or from the standard thermohaline density calculations oceanographers use at ship stations. It was present, in instrument readings from multiple countries’ vessels, for 29 years. It was simply never identified as a current.

What the Keldysh Current Is — and Where It Comes From

To understand how the current forms, it helps to understand the Kara Sea’s position in the Arctic’s plumbing system. The sea sits north of Siberia, between the island archipelago of Novaya Zemlya to the west and Severnaya Zemlya to the east. It is one of the Arctic’s most climatically sensitive seas, with sea-ice coverage declining sharply over recent decades and shipping traffic along the Northern Sea Route — the commercial corridor running along Russia’s Arctic coast from the Kara Sea to the Bering Strait — growing substantially in recent years.

Two branches of warm, salty Atlantic water dominate the region. The Barents Sea branch water (BSBW) crosses the Barents Sea from southwest to northeast, cooling as it goes. Under the standard model, the entire volume of BSBW cools enough to increase in density and sink between Novaya Zemlya and Franz Joseph Land, losing contact with the atmosphere and continuing as a subsurface flow. That picture, Osadchiev and colleagues show, is incomplete.

The northern periphery of BSBW — the outermost edge of the Atlantic inflow, where it has mixed with fresher Arctic surface water — is less dense than the main body. When it cools alongside the main flow, it does not sink. It stays at the surface. As the main body of BSBW descends and is then further sealed off by a layer of cold dense water (CDW) formed through sea-ice production near Franz Joseph Land, that peripheral fraction becomes fully isolated from it — yet it continues to follow the same broad northward and then eastward trajectory, dragged along by the dynamic pull of the vigorous deep flow below.

The result is a surface current shadowing the path of a subsurface current it can no longer touch: warmer and saltier than the surrounding Arctic surface water, tracing the eastern flank of the St. Anna Trough northward to the continental slope, then turning eastward toward the Laptev Sea, while the BSBW proper flows hundreds of meters below.

Caught in Three Decades of Data

The team’s formal characterization of the current relied on 25 hydrographic transects collected across the northeastern Barents Sea and northern Kara Sea between 1995 and 2024 — data gathered by a dozen different research vessels across multiple countries. All measurements used CTD profilers (conductivity-temperature-depth instruments) that record the thermohaline structure of the water column from surface to seafloor. In retrospect, the Keldysh Current appears in all of them as a persistent surface layer noticeably warmer and saltier than the Arctic surface water on either side of it.

A preliminary identification of the current was published by some of the same authors in a 2022 St. Anna Trough paper, based on measurements from two 2021 cruises. The new study expands that initial observation into a full characterization using the 29-year observational record.

The estimated transport — approximately 0.4 Sv — is based on the current’s observed width (~50 km, or 31 miles), depth (~50 m, or 164 ft), and a velocity estimate of roughly 0.17 meters per second (0.56 feet per second). That velocity was itself derived from an indirect measurement: satellite sea-surface salinity observations in August-September 2015, when the Keldysh Current captured the massive Ob-Yenisei river plume and carried it northward through the St. Anna Trough at a trackable rate.

How Does Ocean Science Miss a Current This Size for 30 Years?

Answering that question requires understanding what “measuring” a region means in practice. Oceanographic surveys are expensive, ship time is limited, and the data from any given cruise reflects the positions of a finite set of instrument drops along planned transect lines. Scientists analyzing those measurements have to decide what they are looking at — and for three decades, the warm, salty surface anomaly that appears on the eastern side of the St. Anna Trough was either attributed to local effects or simply not examined for what it would imply about regional circulation.

More fundamentally, the paper acknowledges that its identification of the Keldysh Current rests on thermohaline signature rather than direct velocity measurement. As the authors confirm in the Discussion section, the current is non-geostrophic, meaning its presence could not be reconstructed from satellite altimetry or from the classical density-pressure balance using in situ thermohaline measurements. What that means in plain terms: the two standard methods by which oceanographers infer circulation — altimetry from space, and density-gradient analysis from ship data — are both blind to this class of current. The Keldysh Current is identifiable in the existing data only because the researchers knew to look for a persistent thermohaline anomaly along a consistent pathway over multiple years, and because the dataset was now large enough to make that pattern unambiguous.

The authors note that a full kinematic description of the current — its actual velocity field, its variability, its detailed pathway beyond the Kara Sea continental slope — awaits dedicated mooring deployments and acoustic Doppler current profiler (ADCP) sections. The study’s finding is structural: a current exists; it has a specific formation mechanism and pathway; it was not described in prior literature despite being inadvertently measured for nearly three decades.

A Freshwater Twist

One finding that underscores the current’s broader significance involves its interaction with the Ob-Yenisei plume — the largest river discharge system emptying into the Kara Sea, fed primarily by Siberia’s Ob and Yenisei rivers, which together account for roughly one-quarter of Arctic river discharge.

In certain years, wind forcing pushes the plume northward into the St. Anna Trough, where it encounters the Keldysh Current. When that happens, the current captures the plume and advects it rapidly northward toward the continental slope and the deep Arctic basin. The 2015 event was the clearest documented instance: the plume moved approximately 300 km (186 miles) northward through the trough in the first twenty days of August alone, and by the time in situ measurements were taken on September 25–26, a substantial isolated segment remained in the trough’s northern reaches, as documented in satellite salinity records from August-September 2015.

The volume of freshwater involved in that single event was estimated at approximately 250 km³ (60 cubic miles) — equivalent to about 17 percent of the Kara Sea’s total annual river discharge. What had previously been assumed to spread across the central Kara Sea was instead shunted toward the deep Arctic basin by a surface current no one had named. The study notes that a similar event was observed in September 2024 using updated satellite salinity data.

Why This Changes the Arctic Warming Picture

The Barents and Kara seas are at the center of what oceanographers call “Atlantification” — the progressive warming and salinity increase of the Arctic Ocean driven by intensifying inflows of Atlantic water. The process involves both subsurface Atlantic water shoaling toward the surface and direct surface-layer heat transport. The Keldysh Current now occupies the second category: a direct, surface pathway for Atlantic heat into the Kara Sea and toward the Laptev Sea that was entirely absent from prior circulation models, a gap relevant to Arctic Atlantification as documented by NOAA and in the study’s own conclusions.

The implications for sea-ice forecasting are concrete. The study documents a recurring strip of reduced sea-ice concentration and thickness along the Keldysh Current’s pathway in the northern Kara Sea — delayed freeze-up in autumn, thinner ice in winter and spring, and earlier melt in summer. Prior studies attributed this anomaly to upward heat flux from the deep Atlantic water layer below; the new paper argues the direct surface warming from the Keldysh Current is the more likely explanation, because the anomaly appears specifically along the eastern flank of the St. Anna Trough, not along the western flank where the Fram Strait branch water — another Atlantic inflow — also travels.

The paper also raises an implication for Arctic plastic pollution research. Prior studies assumed the Barents Sea functions as a dead end for surface-floating debris that enters from the North Atlantic. The Keldysh Current creates a direct surface-transport pathway from the Barents Sea through the Kara Sea toward the deep Arctic basin, potentially revising that picture, according to the study’s Discussion section.

The Kara Sea is not an unexplored frontier. It is one of the most studied Arctic seas precisely because of its environmental sensitivity and its position along the Northern Sea Route. The Keldysh Current’s emergence from three decades of existing data is a reminder that what instrumentation can detect shapes what science finds — and that the most consequential unknowns may already be hiding in data already collected, waiting for someone to look at them the right way.

Frequently Asked Questions

What is the Keldysh Current, and why was it only named now?

The Keldysh Current is a surface ocean current in the Kara Sea that carries warm, salty Atlantic-origin water from the northern shore of Novaya Zemlya northward through the St. Anna Trough and eastward along the continental slope. It was formally described in a September 2026 paper in Frontiers in Marine Science, published at doi:10.3389/fmars.2026.1901878. Despite being measurable in data going back to 1995, it was not identified as a current earlier because it is non-geostrophic — it does not create the pressure-gradient signature that satellite altimetry and conventional density-based analysis use to infer circulation. Without that signature, a coherent, named current appears in instrument readings only as an anomalous patch of warmer, saltier water. It took a 29-year multi-vessel dataset and an analysis specifically looking for that pattern to establish that the anomaly was a consistent, coherent flow.

How do scientists normally detect ocean currents, and why didn’t those methods work here?

Standard oceanographic detection relies on two complementary approaches. Satellite altimetry measures sea surface height from space; because a geostrophic current is in balance between a pressure gradient and the Coriolis force, it raises or lowers the sea surface measurably, and that height signal can be inverted to infer the current, as explained in standard descriptions of geostrophic balance. The second approach uses vertical profiles of seawater density collected from ships; currents in geostrophic balance produce systematic density variations across their width, which allow oceanographers to calculate velocity. The Keldysh Current is non-geostrophic — it is driven by dynamic entrainment from the subsurface Barents Sea branch water below it, not by its own pressure field. It therefore raises no sea surface height signal and produces no systematic density anomaly that standard geostrophic calculations would recognize as a current. It was effectively outside the detection range of two of the discipline’s most powerful tools.

What does the Keldysh Current mean for climate change projections of Arctic warming?

The discovery matters for climate modeling in a specific way. Climate models project how Atlantic heat enters the Arctic — both through subsurface Atlantic water inflows and through surface-layer pathways. Until now, no surface-layer Atlantic current was known to exist in this part of the Kara Sea. The Keldysh Current transports approximately 0.4 Sv — about twice the Amazon River’s peak discharge — of warm Atlantic water directly at the surface. That heat is now documented to reduce sea-ice thickness and delay freeze-up along the eastern flank of the St. Anna Trough, an effect previously attributed to subsurface heat flux. Models that lack this pathway may be underestimating the rate of surface-layer Atlantification in the Kara Sea, a concern that maps directly onto Atlantification trends in the NOAA 2025 Arctic Report Card and the study’s conclusions.

Could other unnamed currents be hiding in existing oceanographic data?

The Keldysh Current’s discovery suggests the answer is yes — particularly for currents that are non-geostrophic and therefore invisible to satellite altimetry. The key conditions that allowed the Keldysh Current to go undetected were: a formation mechanism that produces no pressure-gradient signature; a pathway driven by entrainment from a subsurface flow rather than by its own dynamics; and a thermohaline signature — warmer, saltier surface water — that could be and was measured repeatedly, just not identified as a current. Any region where similar conditions exist — a vigorous subsurface flow with a lightly modified surface fraction riding above it — could in principle host a similar undescribed current. The Arctic, with its complex layering of Atlantic, Arctic, and riverine water masses, is a plausible candidate for additional discoveries of this type.