Deep in Norway's Masfjorden, nearly half a kilometer below the surface, something was slowly suffocating. Between 2010 and 2021, dissolved oxygen concentrations crept downward year after year in the isolated basin at the fjord's bottom. By the late 2010s, fish farms had to shut down and relocate. The water had become so depleted of oxygen that marine life could barely survive.

Then, in April 2021, the fjord took its first deep breath in more than a decade.

Scientists monitoring the fjord captured this rare event—a deep-water renewal that flushed the stagnant basin with cold, dense, oxygen-rich water from the coast. The observations, combined with high-resolution computer modeling, reveal not just how these renewal events work, but why they're happening less frequently as oceans warm. Understanding this pattern matters: similar fjords worldwide face the same suffocation risk, with consequences for ecosystems, fisheries, and the millions of people who depend on coastal waters.

The Fjord Trap

Masfjorden stretches 25 kilometers inland from Norway's western coast, a narrow ribbon of seawater rarely wider than two kilometers. Like many Norwegian fjords, it's a spectacular legacy of Ice Age glaciers that carved deep trenches through bedrock. At its deepest point, the seafloor lies 480 meters down.

But what makes Masfjorden geologically interesting also makes it vulnerable. A shallow sill—essentially an underwater ridge—guards its entrance, rising to within 70 meters of the surface. This sill acts like a dam, trapping water below it inside the fjord basin. Surface waters can flow freely in and out, but the deep water sits isolated, cut off from the coastal ocean's circulation.

In this isolation, life consumes oxygen faster than physical processes can replace it. Organic matter sinking from surface waters decays, bacteria respire, fish breathe. Gradually, inexorably, oxygen concentrations decline. Hypoxia develops when oxygen drops below about 61 micromoles per kilogram of seawater. In severe cases, anoxia—complete oxygen depletion—can occur.

The basin becomes inhospitable. Growth slows, survival rates plummet, behavior changes, immune systems weaken. For commercial fish farms operating in Norwegian fjords, extended stagnation spells economic disaster.

Stagnation ends only when a deep-water renewal occurs: when coastal water denser than the trapped basin water manages to spill over the sill and sink to the bottom, displacing the old, oxygen-depleted water upward and outward. The basin gets flushed with fresh water. Oxygen levels recover. The biological clock resets.

These renewals aren't predictable. Some fjords experience them every few weeks. Others wait years. Masfjorden historically renewed every five to ten years, but observations between 1990 and 2020 suggested the interval was lengthening.

Catching the Event

Torunn Sagen and her colleagues from the University of Bergen had been monitoring Masfjorden since September 2019 with a combination of moorings measuring temperature, salinity, oxygen, and currents, plus regular ship-based sampling along the fjord's length. By early 2021, the bottom water held only about 100 micromoles of oxygen per kilogram—not yet hypoxic, but heading that direction.

On April 7, 2021, instruments at the sill detected something unusual. Water density at 68 meters depth exceeded the maximum density measured in the deep basin. Dense water was arriving at the sill.

The high-density episode lasted twelve days. A second pulse arrived on April 23 and persisted through May 1. These weren't brief fluctuations—they represented sustained delivery of water heavy enough to sink through the fjord basin.

The team estimated that during these episodes, dense water was flowing over the sill at 0.08 to 0.15 meters per second. Fast enough to completely replace all the water below the 70-meter sill depth in 15 to 27 days.

Eight kilometers into the fjord, in the deep inner basin, the moorings told the story of the renewal's arrival. Bottom temperatures had been stable since early March. Then, on April 16, temperature at 370 meters began rising gradually. On April 28, twenty days after the first high-density episode began at the sill, temperature at 472 meters suddenly jumped 0.15 degrees Celsius. Current meters registered flowing water where normally they detected only sluggish drift.

By August, when the research team returned for their next survey, the transformation was complete. The two-layered structure they'd observed in April—an upper layer from 100 to 300 meters and a lower layer below—had vanished. Density in the upper layer had increased by 0.06 kilograms per cubic meter. Most dramatically, dissolved oxygen below 300 meters had climbed by more than 70 micromoles per kilogram.

The fjord had renewed itself. The question was: how?

Wind, Water, and Upwelling

The renewal mechanism involves a precisely orchestrated sequence of atmospheric and oceanic processes. It begins not in the fjord but out on the open coast, where winds determine which water masses sit at which depths.

Along Norway's coast, winds from the south tend to push surface water toward shore, a process called downwelling. This forces coastal water to sink, deepening the interface between light surface water and the denser water below. Northerly winds do the opposite. They push surface water away from shore, and deeper water rises to replace it—coastal upwelling.

For deep-water renewal, upwelling matters crucially. The dense water capable of renewing fjord basins typically sits below the sill depth at the coast. Northerly winds must be strong enough and sustained enough to lift this dense water up to sill level, where it can flow into the fjord.

Sagen's team analyzed wind data from Fedje, a meteorological station outside the Masfjorden system. Throughout April 2021, several northerly wind events struck the coast, some reaching 10 to 15 meters per second and lasting two to five days.

But were these winds sufficient? To quantify this, the researchers adapted a theoretical framework involving what's called the wind impulse—essentially the integrated wind stress over time. They calculated a critical threshold: the wind impulse needed to lift water with density matching the maximum fjord basin density up to sill depth.

The observations showed that wind impulse exceeded the critical threshold on April 5-6 and again April 21-25. These dates matched precisely with the high-density episodes observed at the sill two days later.

The mechanism emerged clearly. Northerly winds drove coastal upwelling. Dense water rose from depth at the coast, reached sill level, and began flowing into the fjord as a bottom current, following the bathymetry downward into the basin.

The Model's View

One challenge in studying deep-water renewals is their episodic, unpredictable nature. Relying solely on observations requires extensive, long-term monitoring programs covering many fjords—expensive and logistically complex.

High-resolution numerical models offer an alternative. If a model accurately reproduces the coastal processes that trigger renewals, it might predict when conditions favor renewal events, even in fjords without continuous monitoring.

The research team tested Norfjords, a regional ocean model with 160-meter horizontal resolution covering Norwegian fjords. They compared its performance against their Masfjorden observations.

The model captured the essential dynamics. It reproduced coastal upwelling in response to northerly winds with realistic timing. It showed dense water propagating from the coast along Fensfjorden's southern border toward Masfjorden's sill. It generated a high-density episode between April 6-17, coinciding with the first observed episode.

The model produced a renewal event in the fjord basin starting one day after its modeled high-density episode began. However, the renewal was partial, affecting water at 150 meters depth but not reaching the true bottom at 250 meters. The model underestimated water density both at the coast and in the fjord basin.

Despite these limitations, the model demonstrated something important: it could successfully describe the coastal upwelling process and advection of dense water from coast to sill. The deficit lay not in representing the wind-driven coastal dynamics but in accurately capturing the small density differences that determine whether renewal occurs.

A Delayed Response

The observations revealed an intriguing puzzle. The first high-density episode started April 8. Dense water should have reached the deep inner basin in under two days based on simple advection timescales. Yet instruments at 370 meters didn't register changes until April 16, and the bottom at 472 meters not until April 28—eight and twenty days after the episode began.

The researchers propose that the renewal unfolded in two phases, shaped by Masfjorden's two-sill, two-basin structure. The outer basin, 300 meters deep, sits behind the main 70-meter sill. A second sill at 180 meters separates it from the deeper inner basin.

Dense water flowing over the main sill first had to fill the outer basin to the 180-meter sill depth before it could spill into the inner basin. More significantly, turbulent mixing during the inflow—as dense bottom currents entrained lighter water from above—reduced the density of inflowing water as it progressed into the fjord.

Initially, the inflowing water was only slightly denser than the maximum basin density—less than 0.1 kilograms per cubic meter difference. Mixing likely reduced its density below the basin maximum, causing it to level out in the upper layer rather than sinking to the bottom. Only as the high-density episodes continued did the inflowing water become dense enough to penetrate to the basin floor.

Evidence for this two-phase renewal appears in the August oxygen profile, which shows a local minimum around 275 meters—a remnant of "old" oxygen-depleted water caught between the renewed upper and lower layers. The oxygen profile suggests the inflowing water first spread through the mid-depth layer before eventually sinking and filling the basin from below.

The Changing Frequency

Deep-water renewals in Masfjorden have been monitored, albeit intermittently, since the 1970s. The historical record shows a troubling trend: renewals are happening less often.

Before 1990, Masfjorden typically renewed every two to five years. After 1990, the average stagnation period stretched to five to ten years. Between 1990 and 2020, observations suggest only five major renewal events occurred.

The culprit appears to be coastal ocean warming and freshening. As global oceans absorb heat from climate change, coastal waters warm. Warmer water is less dense. Simultaneously, increased precipitation and glacier melt add freshwater to coastal waters, further reducing density.

These changes show up clearly in the data from Sognesjøen, a coastal monitoring station operated since 1935. Annual maximum density at 75 meters—the depth most relevant for Masfjorden's renewal—has declined since 1990.

For renewal to occur, coastal water density must exceed basin water density. As coastal maximum densities decline while basin densities decrease only slowly through diffusive processes, the conditions permitting renewal become rarer.

The 2021 renewal succeeded because northerly winds strong enough to drive significant upwelling coincided with a period when the basin density had decreased through eleven years of diffusive mixing. The critical threshold became just barely achievable.

Models suggest this pattern extends beyond Masfjorden. A recent study of 101 Norwegian fjords found that deep-water renewals occurred less frequently after 1990 in 65 of them, linked to increased coastal stratification driven by warming and freshening.

Implications and Predictions

The findings matter for practical reasons beyond scientific curiosity. Norway hosts extensive aquaculture operations in fjords, producing salmon and other species in open net pens. Placing farms in poorly ventilated fjords creates problems: biological waste from fish farms increases oxygen consumption in basin waters, potentially worsening hypoxia during long stagnation periods.

Understanding individual fjord renewal frequencies helps aquaculture industries and policymakers identify suitable locations. Fjords with frequent renewals can support more intensive operations. Those prone to long stagnation periods might require restrictions or alternative approaches.

The modeling results, despite their limitations, suggest a path forward. While current regional models may not perfectly capture the small density differences that determine whether specific renewal events occur, they successfully represent the coastal upwelling processes that deliver dense water to sill depths.

This capability could be combined with local fjord monitoring to create a prediction framework. Coastal models identify when upwelling has lifted sufficiently dense water to sill level. Basin monitoring tracks the actual fjord density. Together, they determine when renewal becomes possible.

The modified wind impulse concept provides a quantitative tool: when wind impulse exceeds the critical threshold based on current coastal stratification and basin density, conditions favor renewal. With weather forecasts extending several days, this approach might predict renewal events before they occur, giving aquaculture operators and environmental managers advance warning.

The Broader Picture

Masfjorden's 2021 renewal, captured in unprecedented detail by moorings and ship-based sampling, illustrates processes playing out in semi-enclosed coastal waters worldwide. Fjords in Sweden, Scotland, Canada, Chile, and New Zealand all experience similar dynamics of stagnation and renewal.

As oceans warm and stratification increases globally, the balance shifts. Stagnation periods lengthen. Hypoxia becomes more common and severe. Ecosystems that evolved with periodic renewal on relatively short timescales must adapt to longer intervals between oxygenation events—or fail to adapt.

The observations from Masfjorden offer both warning and opportunity. The warning: what was once a five-year stagnation period stretched to eleven years, with near-hypoxic conditions developing by the end. Climate change is already affecting fjord ventilation in measurable ways.

The opportunity: with sufficient monitoring and modeling, these events become less mysterious. The combination of coastal observations, fjord basin measurements, and high-resolution models can identify the processes controlling renewal frequency. This understanding might inform management strategies, from aquaculture siting to potential interventions.

Some researchers have even proposed artificial ventilation schemes for fjords at highest risk—essentially engineering deep-water renewals through freshwater injection or mechanical mixing. Whether such interventions prove feasible or advisable remains debatable. But as natural renewal frequencies decline, the question shifts from "Can we?" to "Should we?"

For now, Masfjorden has reset its clock. Oxygen concentrations have recovered. The ecosystem has gained a reprieve. But the diffusive processes that gradually reduce basin density have already resumed. The next eleven years are counting down, and absent stronger winds or denser coastal water, the interval might stretch even longer.

The fjord will eventually breathe again. The question is when—and whether the life within it can wait that long.

Publication Details: Year of Publication: 2025; Journal: Estuarine, Coastal and Shelf Science; Publisher: Elsevier Ltd.; DOI Link: https://doi.org/10.1016/j.ecss.2025.109287

Credit & Disclaimer: This article is based on research published in Estuarine, Coastal and Shelf Science. Readers seeking comprehensive technical details—including mooring specifications, model parameters, wind impulse calculations, and complete hydrographic data—should consult the original research paper. This article simplifies complex oceanographic concepts for general readers while maintaining scientific accuracy. Access the full publication and associated datasets at the DOI link above.