Antarctica’s ice loss is locked in, and extra snowfall won’t save it – that’s bad news for low-lying coastal areas elsewhere

The Antarctic ice sheet’s future depends how a chain of events plays out. A new study finds the evidence points in one direction.

Ads

A glacier front meets the sea in Antarctica. The ice sheet holds enough water to raise global sea level by about 190 feet (58 meters). Qiyue SunSea-level rise is one of the most visible signs of climate change, one that is threatening low-lying coastal communities around the world. Today, about 1 billion people live in coastal areas; about 100 million live within 1 meter (3.3 feet) of sea level, highly exposed to coastal flooding.

Over the long term, when it comes to future sea-level rise, one place matters more than any other: Antarctica.

The frozen continent holds enough ice to raise global sea level by about 190 feet (58 meters). Yet scientists have disagreed on a fundamental question: Will Antarctica gain more ice than it loses this century or lose more ice than it gains?

In a new study published in the scientific journal Nature Geoscience, we looked at hundreds of thousands of potential futures and the impact of each on the ice sheet.

We found that Antarctica’s ice loss this century is locked in – even if the world limits global warming to 1.5 degrees Celsius (2.7 degrees Fahrenheit), the 2015 international Paris climate agreement’s most ambitious goal.

Why the Antarctic ice sheet is so hard to predict

The Antarctic ice sheet’s future depends on a chain of events.

Greenhouse gas emissions from vehicles, coal- and natural gas-fueled power plants and other sources around the world determine how much the ocean and air around Antarctica warms. That warming shapes how ice shelves around the continent’s margins melt, and how quickly the glaciers behind them slide over the bedrock. When those ice shelves thin or collapse, studies show the glacier flow speeds up.

A European research team explains ways Antarctica is vulnerable to ice loss from the ocean and how models work. TiPACCS.

Each step relies on assumptions that scientists have yet to agree on, such as how ice shelves collapse. Different assumptions give different answers, and because each step feeds the next, the differences grow – what scientists call cascading uncertainty. As a result, research teams around the world can’t agree on whether Antarctica will gain or lose ice this century.

Snow is part of the puzzle. Warmer air holds more moisture, so as temperatures rise, more snow will likely fall on Antarctica. In some projections, that extra snow outweighs the ice lost to the warming ocean, and the ice sheet actually grows.

Auditioning possible futures

A big reason for this persistent disagreement involves the cost of computing.

Each ice-sheet simulation runs with just one set of assumptions for every step in the chain, from how much greenhouse gas we emit to how the ice behaves. One run can keep a supercomputer busy for days, so no team can test every plausible combination.

Our team of data scientists, glaciologists and oceanographers took a shortcut. An international project called the Ice Sheet Model Intercomparison Project, or ISMIP6, has built a large archive of Antarctic simulations showing how the ice responds under different circumstances. We trained a physics-informed machine learning model on that archive to learn how each simulation’s outcome depends on its physical assumptions, such as how easily ice slides over bedrock. The machine learning model does in a fraction of a second what took the original models days, predicting sea-level change with similar accuracy.

NASA’s Grace satellite has tracked ice mass loss over the years.
NASA

That speed let us run millions of combinations and check hundreds of thousands of possible futures against data from satellites that weigh the ice sheet from space by sensing tiny changes in Earth’s gravity. Only scenarios that were able to match what the satellites have seen since 2002 made the cut.

The scenarios that did make the cut point one direction: Antarctica will lose more ice than it gains this century, even if humanity is able to limit global warming to 1.5 C.

That means Antarctica will contribute more to sea-level rise that’s already being driven by melting of the Greenland ice sheet and mountain glaciers and from warming ocean temperatures that cause water to expand.

What a clearer forecast can do

The biggest lessons from our study are that greenhouse gas emissions matter and extra snowfall won’t save Antarctica. Every ton of greenhouse gas that humans can avoid releasing today reduces how much ice Antarctica loses this century, and the benefit – slower and less sea-level rise – lasts for hundreds of years.

In a scenario with very high emissions, our results show a potential faster domino effect. A fast-warming ocean thins the ice shelves, weakened shelves break apart, and glacier ice flows into the ocean faster. Each step makes the next one easier. Together, these steps could add up to 10 inches (25 centimeters) of sea-level rise from Antarctica alone by 2100 – enough to permanently flood the homes of more than 10 million people.

What different types of models help scientists understand about ice melt, and how they are connected to help predict the Antarctic ice sheet’s future. On the left, how greenhouse gas emissions change the climate, including sea ice, ocean and atmospheric dynamics, which influence one another and levels of global warming. On the right, what goes into an ice sheet model, including ice shelf melt and collapse and how the ice slides. Every stage adds uncertainty.
Yucheng Lin

Of more than 20 physical assumptions about Antarctica, three matter most for understanding future sea-level rise: how ice shelves respond to a warming ocean, how ice slides over bedrock, and how fast the ocean and air around Antarctica warm. Gaining a better understanding of these processes would do the most to sharpen sea-level projections.

What’s still missing

Our study is far from the final word.

Some processes are missing from today’s simulations, so our machine learning model can’t learn about them – for example, how ice cracks and breaks, and how rivers and lakes beneath the ice sheet change over time. Any of these could change our understanding of how quickly Antarctica’s ice reaches the sea and whether coastal communities have enough time to adapt.

What our study does offer is a much faster way to test combinations of assumptions that are too costly to simulate one by one. That provides a clearer basis for the decisions coastal communities around the world are already having to make.
Yucheng Lin receives funding from City University of Hong Kong
(project number 9610793). Robert Kopp receives funding from the U.S. National Science Foundation (NSF) and the National Aeronautics and Space Administration (NASA).

Tags: Academic Journalism, The Conversation, The Conversation Academic News