Stopping Antarctica from melting entirely isn’t realistic at current warming levels, but a combination of aggressive emissions cuts and experimental engineering projects could slow the loss dramatically. Antarctica is shedding roughly 135 billion tons of ice per year, raising global sea levels by about 0.4 millimeters annually. That rate is accelerating, and some of the most vulnerable glaciers may already be approaching tipping points. The strategies being developed range from planetary-scale carbon reduction to placing physical barriers on the ocean floor.
Why Antarctica Is Melting From Below
Most people picture ice melting from warm air above, but the primary driver of Antarctic ice loss is warm ocean water circulating beneath floating ice shelves. A deep current called Circumpolar Deep Water pushes relatively warm, salty water along the seabed toward the base of glaciers, eating away at the ice from underneath. As ice shelves thin, they lose their ability to hold back the massive glaciers behind them, which then flow faster into the sea.
This process is especially dangerous at Thwaites and Pine Island glaciers in West Antarctica. Both sit on bedrock that slopes downward inland, meaning that as the ice retreats, it exposes ever-deeper terrain to warm water, potentially triggering runaway collapse. The West Antarctic Ice Sheet alone holds enough ice to raise global sea levels by over three meters.
Cutting Emissions: The Single Biggest Lever
Every fraction of a degree matters for Antarctica’s ice. Recent research estimates that the West Antarctic Ice Sheet’s collapse threshold sits at around 1.5°C above preindustrial temperatures, a level the world is already brushing against. The marine-based portions of the much larger East Antarctic Ice Sheet have a higher threshold, estimated at 2 to 3°C, but crossing it would be catastrophic on a scale that dwarfs anything from the west.
Holding warming to the Paris Agreement targets is the only intervention that addresses the root cause. Every other strategy discussed below is a stopgap designed to buy time while the world reduces greenhouse gas emissions. Without deep cuts to carbon pollution, no engineering project can keep pace with the accelerating forces driving ice loss.
Seabed Curtains to Block Warm Water
One of the most detailed engineering proposals involves anchoring enormous flexible curtains to the ocean floor to physically block warm deep water from reaching glacier grounding lines. The curtains are buoyant, stretching upward from the seabed like underwater sails. As tides and currents shift, each panel passively adjusts its lean angle, balancing ocean forces against its own buoyancy without rigid supports.
A feasibility study published in PNAS Nexus modeled an 80-kilometer curtain installed in waters over 600 meters deep that could help stabilize both Pine Island and Thwaites glaciers for centuries. The optimal curtain depth would keep the top of the barrier below about 500 meters, deep enough that most icebergs (which typically extend 300 to 500 meters below the surface) would pass overhead without snagging it. The highest-leverage section of the route is less than 5 kilometers long, though the full protective line would stretch much farther.
The estimated price tag is $40 to $80 billion for construction over roughly a decade, plus $1 to $2 billion per year in maintenance. That sounds enormous, but the comparison is striking: if Thwaites and Pine Island collapse, the resulting sea level rise would require an estimated $40 billion per year in coastal sea dike investment worldwide. Spending a fraction of that to prevent the problem is, at least on paper, a bargain.
Artificial Sills on the Seafloor
A related concept involves building underwater berms or sills at strategic points where warm water funnels toward glacier bases. These raised mounds on the seafloor would physically obstruct warm currents, much like a speed bump redirecting water flow. Modeling work on Thwaites Glacier found that blocking warm water access causes ice shelves to thicken, increases their buttressing strength, and can actually push the grounding line back toward the sea.
In some simulations, a grounding line that had retreated over 100 kilometers behind its current position was able to recover after a sill was placed. Even a structurally weak sill could delay collapse for centuries by allowing the ice shelf to reground on the barrier itself, creating a self-reinforcing stabilization effect. In Greenland, where warm water is channeled through narrow fjords, artificial sills at these natural choke points could achieve similar results with smaller structures.
Reflecting Sunlight With Atmospheric Particles
Stratospheric aerosol injection is a more speculative approach. The idea is to release reflective particles high in the atmosphere to bounce some sunlight back into space before it warms the surface. NOAA-funded researchers modeled eleven different injection strategies, varying the location and quantity of particles, to see which would best protect Antarctic ice.
The results highlighted a crucial detail: where you inject matters enormously. Concentrating particles in the southern hemisphere gave the best results for slowing Antarctic melt. A simulation using a single injection point in the northern hemisphere actually increased Antarctic melting by warming the surrounding Southern Ocean. Strategies using injections at multiple latitudes showed some ability to reduce ice loss, but none eliminated it entirely.
This approach remains controversial. It doesn’t address ocean acidification or other consequences of high carbon dioxide levels, and stopping injections abruptly could cause rapid rebound warming. It’s best understood as a potential complement to emissions reductions, not a replacement.
Pumping Seawater Back Onto the Ice Sheet
One of the more audacious proposals involves pumping ocean water inland and depositing it as artificial snow on East Antarctica’s surface, essentially returning meltwater to cold storage. Researchers at Columbia University calculated what this would actually take, and the numbers are staggering.
To offset roughly 3 millimeters of sea level rise per year (close to the current global rate), you’d need 90 of the largest pump stations currently under construction in New Orleans, each moving about 360 cubic meters of water per second. The water would need to travel roughly 700 kilometers inland to reach terrain cold and stable enough to keep it frozen for a thousand years, and it would only work on East Antarctica because the western ice sheet is too unstable. The seawater would also need to be converted to snow before being deposited, since liquid saltwater on the surface would accelerate melting.
The energy required just for pumping would exceed 7 percent of the world’s entire current energy supply. Meeting that demand with renewable power would take an estimated 850,000 wind turbines running at full capacity. This concept remains firmly theoretical, useful mainly for illustrating the sheer scale of the problem.
Ecological Risks of Ocean Interventions
Blocking warm water from glaciers doesn’t just affect ice. That same warm, salty deep water carries nutrients that sustain entire marine food webs. Research published in AGU Advances used Greenland’s largest marine-terminating glacier as a case study and found that installing an artificial barrier would likely shut down a natural nutrient delivery system. Meltwater plumes at glacier fronts pull deep, nutrient-rich water upward, feeding massive phytoplankton blooms that support major fisheries.
At Greenland’s Sermeq Kujalleq glacier, this process sustains one of the country’s largest inshore fisheries in Disko Bay. An artificial barrier over the fjord’s sill would cut off the saline water mass responsible for both heat inflow and nutrient supply. The researchers concluded that negative impacts on fisheries raise serious questions about social viability, concerns that need to be evaluated alongside technical feasibility rather than as an afterthought. Similar dynamics likely exist around Antarctic glaciers, though the ecosystems there are less well studied.
Who Is Funding This Work
Most Antarctic stabilization research remains in the modeling and early feasibility stage. New Zealand recently announced $49 million to extend its Antarctic Science Platform for another seven years, making it that country’s largest investment in Antarctic research. The platform was originally established to study Antarctica’s behavior in a world that reaches 2°C of warming. NOAA has funded research into solar radiation management modeling, and individual university teams have published feasibility analyses for seabed curtains and artificial sills.
No government or international body has committed to building any of these interventions. The Antarctic Treaty system, which governs activity on the continent, would need to accommodate large-scale engineering projects, a process with no precedent. Coordination between nations on funding, construction, and long-term maintenance remains one of the biggest practical obstacles, potentially harder to solve than the engineering itself.