In February 2021, an underwater robot slipped beneath iceberg A-68A in the Scotia Sea. What it found upends assumptions about how melting giants reshape the ocean around them.

A-68A was colossal. When it broke from Antarctica's Larsen-C ice shelf in 2017, it covered an area larger than Delaware. For years it drifted north, shedding fragments, until it neared South Georgia—a remote island where ocean currents converge and biological productivity explodes.

The glider's mission was simple in concept, perilous in execution. Track the iceberg. Measure temperature, salinity, density. Watch what happens when frozen freshwater meets saltwater in motion.

Two days in, the glider became trapped beneath the ice at 163 meters depth. Twelve hours later, it had risen to 112 meters—evidence of the iceberg's uneven keel scraping through the water column. For weeks, the robot collected data no human instrument had captured before: a real-time portrait of an iceberg rewriting the ocean's vertical structure.

The architecture of cold water

The Southern Ocean has a peculiar anatomy. Surface waters sit atop a cold subsurface layer called winter water—a remnant from the previous season's freeze. Below that lies circumpolar deep water: warmer, saltier, loaded with nutrients that have accumulated over centuries of decomposition and remineralization.

Normally, these layers remain distinct. The winter water acts as a density barrier, a kind of potential energy wall that prevents surface and deep waters from mixing easily. This separation matters. It controls how much heat the ocean can absorb from the atmosphere, how much carbon dioxide gets sequestered in the deep, and whether phytoplankton at the surface can access the nutrients they need to bloom.

A-68A disrupted all of it.

As the iceberg's base melted, it generated turbulence. Cold, fresh meltwater mixed with ambient seawater and became buoyant. That mixture rose, carrying with it water from as deep as 238 meters—well below the iceberg's estimated 139-meter draft. The glider measured intrusions: pockets of water warmer and saltier than they should have been at those depths, clear signatures of upwelling driven by the iceberg's passage.

Quantifying this effect required isolating the meltwater signal from the region's chaotic background. The Scotia Sea is a tangle of fronts, eddies, and meanders. Historical hydrographic data helped establish a baseline—what the ocean looks like without an iceberg. The glider profiles, sorted by distance from A-68A, revealed three regimes: far (15 kilometers away), near (2.6 kilometers), and adjacent (less than 300 meters).

The differences were stark.

Two peaks, one erosion

Adjacent to the iceberg, a fresh cold cap sat at the surface—roughly nine meters deep. This pushed warmer, saltier surface water downward, creating a second peak in stratification at around 44 meters depth. Meanwhile, the winter water layer had been eroded. Its characteristic temperature minimum, usually a broad stabilizing feature, had nearly vanished.

Farther from the iceberg, the profile began to resemble the historical norm. But it wasn't identical. The surface remained warmer and saltier than climatology predicted. Below, the water was cooler and fresher. The winter water stratification, once eroded, had partially reformed—but shallower and stronger than before.

The researchers used a method called Gade line analysis to quantify basal meltwater content. Each intrusion in temperature-salinity space traced back to a source depth in the permanent thermocline. By determining the relative proportions of ambient water and meltwater along density surfaces, they calculated an average meltwater contribution of 0.52 cubic meters per square meter, distributed over a vertical extent of 106 meters.

Scaling this up: assuming the influence extends three kilometers from the iceberg's edge, the basal melt flux ranged between 690 million and 1.4 billion cubic meters per day, depending on whether the iceberg's velocity or the geostrophic flow speed dominated advection. Satellite altimetry, which tracked changes in the iceberg's thickness over time, estimated 1.7 billion cubic meters per day.

The agreement is remarkable given the uncertainties. Satellite measurements detect meltwater when it's created. Oceanographic measurements detect it when it's released and mixed into the water column. The glider was also trapped on the iceberg's shallower side, complicating assumptions about uniform melt rates around the perimeter.

Turbulence at the boundaries

To confirm that turbulence was redistributing meltwater vertically, the team calculated the Richardson number from ship-based measurements. This dimensionless ratio compares stratification (potential energy) to shear (kinetic energy). When it drops below 0.25, turbulent mixing likely occurs.

The ship profiles, taken 2 and 4.5 kilometers from A-68A, showed active mixing beneath the fresh cold surface cap, where warmer waters had been pushed down, and near the base of the meltwater intrusions—possibly a boundary layer dragged by the iceberg itself.

This wasn't passive diffusion. The iceberg generated its own circulation: sidewall melt, surface runoff, wake effects, and the buoyant plume from basal melting all contributed. The mixing extended deeper than the iceberg's draft, reaching waters that would otherwise remain isolated.

Biology in flux

The glider carried optical sensors for chlorophyll and backscatter. Close to the iceberg, surface chlorophyll was low while backscatter remained high—a signature of sediment-laden meltwater diluting phytoplankton or reducing light penetration through turbidity.

At 16.7 kilometers from the iceberg, both chlorophyll and backscatter increased. If the iceberg was moving at 0.13 meters per second, that distance translates to roughly 36 hours after passage.

The growth rate implied by integrated chlorophyll changes matched or fell below the maximum for phytoplankton at those temperatures: 0.5 to 1.5 doublings per day. This suggests the biomass increase was local, not advected from elsewhere.

Phytoplankton blooms following iceberg passage have been documented before, often delayed by six to ten days. The mechanisms are complex: nutrient delivery from upwelled deep water and iceberg-hosted sediments, but also dilution by meltwater, changes in light availability due to stratification shifts, and the time required for cells to acclimate and divide.

South Georgia's waters are already productive, yet micronutrient iron often limits growth in the broader Southern Ocean. The upwelling documented here brought nutrient-rich circumpolar deep water—and potentially iron-laden terrigenous material from the iceberg—into shallower layers beneath the freshwater cap.

Whether this fertilization outweighs the dilution and turbidity effects depends on the phytoplankton community structure, the nature of nutrient limitation, and the persistence of stratification changes. The glider data alone can't resolve these questions, but they establish the physical preconditions.

A historical fingerprint

To assess whether this phenomenon is regionally important, the researchers compiled hydrographic profiles from 2004 to 2021 and searched for years when large icebergs transited the Scotia Sea during summer months. Two events stood out: iceberg B-17a in April 2015, and the A-68 fragments in February 2021.

Climatological profiles from these periods showed elevated and shallower stratification maxima compared to the historical median—consistent with the glider observations near A-68A. As separation from the iceberg increased, the maxima deepened and surface waters became fully mixed. Below the winter water core, stratification remained elevated, likely reflecting basal meltwater influence persisting beyond the iceberg's immediate vicinity.

The pattern held across both events despite their geographic and temporal differences, suggesting that giant iceberg passage leaves a consistent physical signature on winter water structure.

Implications for a warming ocean

The Southern Ocean absorbs a disproportionate share of anthropogenic carbon dioxide and heat. One control on this uptake is stratification at the base of the mixed layer. Stronger stratification inhibits vertical exchange, allowing heat and carbon to accumulate near the surface before being subducted along density surfaces.

Recent research shows that global stratification at this interface has increased over the past five decades. Iceberg melting and lateral mixing has been hypothesized as a contributor to winter water formation. This study provides direct observational evidence that giant icebergs increase stratification at the base of the mixed layer in the Southern Ocean.

It also demonstrates that icebergs modify winter water properties outside of winter—an aseasonal mechanism that could alter the temperature and salinity signatures that persist into the following season.

As West Antarctica continues to lose ice mass, the number of deep-drafted icebergs entering the Southern Ocean will likely increase. Individual icebergs vary in draft, speed, trajectory, and sediment load. The cumulative effect on upper-ocean physics, nutrient supply, and biological productivity remains uncertain.

Climate models do not yet represent iceberg-ocean interactions explicitly, largely because field measurements have been scarce and difficult to obtain. The glider deployment beneath A-68A was unprecedented in its proximity and resolution.

What emerges is a picture of icebergs as agents of vertical exchange—machines that erode density barriers, lift deep water toward light, and leave behind a restructured water column. The fresh cold cap at the surface. The eroded winter layer. The warmer, saltier remnant where stratification reforms.

These changes matter for carbon export, ecosystem function, and the ocean's capacity to buffer climate change. Only through direct observation and improved modeling will their full impact be quantified.

A-68A eventually disintegrated. Its largest fragment, A-68D, is gone. But the ocean it passed through retains the memory of that passage—at least until the next winter's freeze rewrites the structure once again.

Credit & Disclaimer: This article is a popular science summary written to make peer-reviewed research accessible to a broad audience. All scientific facts, findings, and conclusions presented here are drawn directly and accurately from the original research paper. Readers are strongly encouraged to consult the full research article for complete data, methodologies, and scientific detail. The article can be accessed through https://doi.org/10.1038/s41561-025-01659-7