Sea levels could rise nearly twice as much as previously predicted by the end of this century if carbon dioxide emissions continue unabated, an outcome that could devastate coastal communities around the globe, according to new research published Wednesday.

The startling findings, published in the journal Nature, paint a far grimmer picture than current consensus predictions, which have suggested that seas could rise by just under a meter at most by the year 2100. Those estimates relied on the notion that expanding ocean waters and the melting of relatively small glaciers would fuel the majority of sea level rise, but did not factor in changes to the massive ice sheets of Greenland and Antarctica.

The scientists behind Wednesday’s study used sophisticated computer models to decipher a longstanding riddle about Antarctica: how did it surrender so much ice during previous warm periods? They found that similar conditions in the future could lead to monumental and irreversible increases in sea levels. If high levels of greenhouse gas emissions continue, they concluded, oceans could rise by close to 2 meters in total (more than 6 feet) by the end of the century. The melting of ice on Antarctica alone could cause seas to rise more than 13 meters (42 feet) by 2500.

The projection “nearly doubles” prior estimates of sea level rise, which had relied on a “minimal contribution from Antarctica,” said Rob DeConto of University of Massachusetts, Amherst, who authored the study with David Pollard of Penn State University.

EXPERTS PRAISE NEW MODEL

The research already has created a buzz in the community of scientists studying Antarctica, and experts largely praised the new model as thorough and impressive, while noting its remaining uncertainties.

“People should not look at this as a futuristic scenario of things that may or may not happen. They should look at it as the tragic story we are following right now,” said Eric Rignot, an expert on Antarctica’s ice sheet and an earth sciences professor at the University of California, Irvine, who was not involved in Wednesday’s study. “We are not there yet … with the current rate of emissions, we are heading that way.”

Should the new research prove correct, it could trigger a “tectonic shift” in expectations for the speed and severity of the sea level problem, said Ben Strauss, director of the program on sea level rise at Climate Central, an independent organization of scientists based in New Jersey. He said that while the study’s findings represent potentially grave problems for many coastal areas in the decades ahead, the century beginning in 2100 could see truly catastrophic shifts, unless societies make sharp cuts in greenhouse gas emissions.

“Under the high emissions scenario, the 22nd century would be the century of hell,” Strauss said. “There would really be an unthinkable level of sea rise. It would erase many major cities and some nations from the map. … That century would become the century of exodus from the coast.”

Places as far flung as South Florida, Bangladesh, Shanghai, Hampton Roads in Virginia and Washington, D.C., could be engulfed by rising waters, Strauss said. Even by 2100, Miami Beach and the Florida Keys could begin to vanish. New Orleans essentially could become an island guarded by levies. Floods that pushed as far inland as the surge from Hurricane Sandy could ravage parts of the East Coast with far greater frequency.

HUMAN BEHAVIOR CAN ALTER OUTCOME

The researchers behind Wednesday’s study make clear that their model has limitations and that human behavior can alter the possible outcomes. For instance, the worst-case scenario – of seas rising nearly 4 feet due to Antarctic ice loss alone by 2100 – assumes that very high emissions continue for carbon dioxide and other greenhouse gases.

In Paris late last year, world leaders forged an historic agreement to begin scaling back such emissions in coming years. They embraced the goal of holding global warming “well below” 2 degrees Celsius above pre-industrial levels, but at the same time, it has been widely noted that current country-level commitments to cut emissions fall far short of this target.

But even under a more moderate emissions scenario, Wednesday’s study found that the Antarctic contribution to sea level rise still could reach about 2 feet by 2100, and much more by 2500. Only if countries sharply reduce emissions does the model show that it’s possible to preserve Antarctica in roughly its current state.

“This research highlights the importance of doing even much better than the Paris agreement if we’re going to save our coastal cities,” Strauss said.

DeConto and Pollard arrived at their projections about future sea level rise by first turning to the past. Their study is based on an improved understanding of two past warm eras in Earth’s history that featured much higher seas, known as the Pliocene and the Eemian. The Pliocene was a warm period about 3 million years ago, when atmospheric carbon dioxide levels are believed to have been about what they are now – 400 parts per million. Sea levels are believed to have been significantly higher than now – perhaps 30 feet or more. The Eemian period, between 130,000 and 115,000 years ago, also featured sea levels 6 to 9 meters above current levels, with global temperatures not much warmer than our current era.

Sea level rise on the scale seen in those eras likely required a loss of ice not just from Greenland, but also from Antarctica. But previous computer models of Antarctica have failed to accurately reproduce such scenarios. Scientists had spent “years of struggling to be able to simulate tens of meters of sea level rise in the Pliocene,” DeConto said. “This has been a longstanding problem for us. And we had known for years that we’re probably missing some important underlying physics.”

TAKING INTO ACCOUNT VAST GLACIERS

Scientists already knew that key parts of Antarctica, and especially West Antarctica, feature a condition called “marine ice sheet instability.” That is, vast glaciers are already rooted below sea level and lie on downward sloping seabeds. Warm water can not only melt them from below, but as the glaciers retreat, more and more ice will be exposed to melting.

The new study factors in not only this process, but two new ice processes that have scientists already have seen destabilize several glaciers in Greenland: “hydrofracture,” in which water formed by the melting of snow and ice atop a glacier’s stabilizing ice shelf causes it to break up; and “cliff collapse,” in which a sheer ice cliff 100 meters or more above sea level becomes unstable and crashes repeatedly into the ocean below. Both phenomena can speed up the pace of ice loss from glaciers and cause sea level rise.

“Build a little sand castle and it is fine; too high and it may break,”said Richard Alley, a glaciologist at Penn State University who has published previously with DeConto and Pollard, describing the revelations regarding ice cliff collapse.

Knut Christianson, a glaciologist at the University of Washington in Seattle, said the new work will spur additional research to determine precisely what happens at glaciers where cliff collapses and so-called “calving” occur. “It’s a more comprehensive analysis than before, and it certainly indicates that we should look more closely to see whether or not the way they treat these processes in the model is accurate in the real world,” he said.

The research further undermines a string of sea level projections from the United Nations’ Intergovernmental Panel on Climate Change, which have been faulted for being too conservative.

In 2013, the body projected that for the same high-end emissions scenario used in the current study, sea level rise by the year 2100 would be between 0.52 and 0.98 meters (1.7 and 3.22 feet), relatively little of which would come from the ice sheets of Greenland and Antarctica. It noted that beyond this likely range, only Antarctica’s marine-based regions could conceivably contribute a lot more, but the panel found that “there is medium confidence that this additional contribution would not exceed several tenths of a meter.”

The new study challenges that reasoning. It also emerges as mounting research has pointed at one region of Antarctica in particular – the Amundsen Sea sector of remote West Antarctica, centered on the enormous, marine-based Thwaites glacier – as particularly vulnerable.

If the projections in Wednesday’s study prove correct, they could present especially bad news for U.S. coasts. The reason is gravity: Antarctica’s enormous mass pulls the ocean toward it, and when it loses significant mass, seas would surge back toward the opposite end of the world.

“Sea level rise is not going to be felt evenly over the surface of the Earth. It’s really bad for New York, Boston. We are sort of in the bullseye,” DeConto said.