According to the findings of a new study published in the journal Nature Geoscience, reducing greenhouse gas emissions now could limit the amount of ice Antarctica loses during this century and reduce the sea-level rise faced by coastal communities.
Reducing Greenhouse Gas Emissions
The study, reported by an international team of scientists that includes Rutgers University climate scientist Robert Kopp, provides evidence that today’s emissions decisions could affect how much water Antarctica adds to the oceans by 2100. This was reported by phys.org.
The team found that the probability that the most ambitious emissions-reduction target, consistent with the goals of the Paris Agreement — the international treaty on climate change adopted in 2015 — would lead to less ice loss in Antarctica by the end of the century than a very high-emissions scenario is at least 89%.
Robert Kopp notes that reducing emissions now can limit the risks future generations will face. However, coastal communities will still need to prepare for sea-level rise by making appropriate decisions to protect critical infrastructure.
Warming Could Weaken Antarctica’s “Ice Brakes”
The possibility of limiting future damage comes with a troubling conclusion: it is highly likely that Antarctica will lose ice overall by the end of this century even if countries sharply reduce emissions. According to the researchers, the probability of this is at least 92% under the most ambitious scenario, in which the world reduces carbon dioxide emissions to net zero around 2050.
Antarctica contains the world’s largest reserves of frozen freshwater. Its enormous ice sheet rests on land, and when this ice enters the ocean, it adds water and raises sea levels. Floating ice shelves along the continent’s edges help hold back the glaciers behind them, acting like brakes.
The researchers studied how these “brakes” could weaken as the climate warms. According to University of Hong Kong researcher Yuchen Lin, warm ocean water melts ice shelves from below, causing them to thin. As ice shelves weaken or collapse, glaciers can flow into the sea more quickly. The rate of this loss is also affected by how easily the ice slides over the surface beneath it. Under certain combinations of conditions, these changes can reinforce one another and accelerate the movement of ice into the ocean.
Estimates of Global Sea-Level Rise
To investigate this possibility, the team analyzed combinations of assumptions that produced the highest projected losses while remaining consistent with satellite observations. Under very high emissions, this selected high-end scenario analysis produced an average estimate of sea-level rise from Antarctic ice loss by 2100 of about 6 inches (15 centimeters), with an upper estimate of about 10 inches (25 centimeters).
These figures describe a possible high-impact scenario rather than the study’s central forecast. The 25-centimeter estimate is not an absolute upper limit. It also reflects only Antarctica’s contribution and does not include water from melting ice in Greenland and mountain glaciers, or the expansion of seawater as it warms.
The researchers found that additional snowfall under warmer conditions is unlikely to fully compensate for Antarctic ice loss. Antarctica responds slowly to climate change, which means the effects of today’s emissions may persist long after they enter the atmosphere.
Artificial Intelligence Helps Refine Forecasts
According to Kopp, reaching clear conclusions about Antarctica has long been a difficult task for researchers. Computer models must connect global warming with changes in the surrounding ocean and atmosphere and then calculate how the ice responds. Different assumptions at each stage can produce dramatically different results.
Running simulations detailed enough to study all of these possibilities would require enormous computing resources. Instead, the team used machine learning — a branch of artificial intelligence — to analyze an existing set of ice-sheet simulations, allowing the researchers to quickly explore different combinations of assumptions.
The machine-learning tool learned the relationships between assumptions in those simulations and their results. It could then estimate outcomes much faster than the original models, enabling the researchers to explore possibilities that would otherwise have been too computationally expensive to test in detail.
They compared the results with satellite measurements of Antarctic ice loss from 2002 to 2021, giving greater weight to combinations that better matched the observations. This comparison narrowed the range of projected outcomes and increased confidence that Antarctica will lose ice overall during this century.
Key Sources of Uncertainty in the Models
The analysis also helped identify where additional research could have the greatest impact. The three main sources of uncertainty are related to how warming ocean water causes ice shelves to melt, how ice slides over bedrock, and how the ocean and atmosphere around Antarctica warm.
According to the scientists, more accurate measurements of these processes could improve forecast precision and help communities make decisions about development and the protection of coastal areas.
However, the researchers noted that their analysis cannot account for processes that were not included in the original simulations, including some interactions between ice, ocean, and atmosphere, as well as changes in water flow beneath the ice sheet. These gaps leave open the possibility that actual outcomes could fall outside the range projected by the study.
