An important part of this effect is driven by the flowing of the Earth's mantle, a layer of thick, oozing rock beneath the Earth's crust, said Michael Bevis, a geophysicist at the Ohio State University and a co-author on the new study. When a heavy weight, such as a huge ice sheet, forms on the Earth's surface, the resulting high pressure causes the rocky mantle to begin flowing out from underneath it. When the weight is removed, the mantle gradually begins to flow back into place.
Because satellite studies generally draw their conclusions about ice loss based on changes in the Earth's surface, scientists must make corrections to account for this effect. But the new paper suggests that our current correction methods may not be entirely accurate - at least, not for Greenland. The study points specifically to the measurements yielded by the Grace satellites, a set of twin crafts that estimate ice loss based on changes in the pull of gravity as they orbit around the earth.
"What Grace measures is mass change, but it can't really tell the difference between ice mass and rock mass," said Bevis. "So you have to give it a model. [And] the model we were using was not quite right."
The study draws its conclusions using data from a network of GPS sensors placed around Greenland, which have helped detect how fast the earth there is springing back up. The researchers were able to use these recent measurements to estimate the rate at which land in Greenland has been moving back into place since the last ice age, which reached its peak more than 25,000 years ago.
When they compared these estimates to some of the models that have previously been used in reconstructions of Greenland's glacial history, they found that the findings didn't match up, leading them to conclude that the models' assumptions about the flow of rock beneath Greenland were incorrect.
These models have typically relied on standard assumptions about the way the Earth's mantle flows in most parts of the world, Bevis noted. But the researchers suggest that millions of years ago, an especially intense region of molten rock beneath the Earth's surface - known as a "hotspot" - changed the consistency of the mantle beneath Greenland, causing it to move in different ways. This hotspot still exists, they've noted, but it has since migrated and currently resides beneath Iceland, where it's historically been responsible for the high levels of volcanic activity in that part of the world.
Without taking the influence of this hotspot into effect, the researchers suggest, previous models of Greenland's behaviour were incorrect.
So they created a modified model, tweaking its assumptions about the mantle so that the results were consistent with their GPS estimates.
Then, they used the modified model to create a reconstruction of Greenland's glacial history.
First, the results suggest that satellite studies have been underestimating the current mass loss in Greenland by about 18 billion tonnes per year.
Bevis pointed out that it's actually a fairly modest correction, suggesting a rate of ice loss that's only about 7.5 per cent faster than previously estimated.
Ian Joughin, a glaciologist at the University of Washington who was not involved with the paper, added that the new findings are still within the uncertainty limits of the results from many previous satellite studies.
But this isn't actually the most important part of the study, Bevis said. The reconstructed history was able to identify not only how much ice has been lost over the last few thousand years, but also where the losses have been coming from.
Over the past two decades, scientists have found that a relatively small set of glaciers in Greenland are responsible for more than 70 per cent of the ice sheet's total losses. The new study finds that these same regions have actually been contributing to a hefty portion - about 40 per cent - of Greenland's ice losses for many thousands of years. And the researchers pointed out that another recent study found similar results over the last century.
According to Bevis, these studies reinforce one another, suggesting that many of the same glaciers have been causing the lion's share of Greenland's ice loss for thousands of years now - a finding he says could be helpful when making predictions about the future of the ice sheet.
However, some experts are more cautious. Joughin noted that some rapidly flowing glaciers may not have much room left in the future to continue retreating, meaning they may not always be the ones contributing the most to Greenland's ice loss.
But the study doesn't just raise questions about Greenland. Bevis and his colleagues have pointed out that some of the same problems could exist with our current estimates of ice loss in Antarctica. The problem there is that Antarctica is so much bigger than Greenland - and although a GPS network exists there as well, the sensors are spaced much farther apart, meaning it may be much more difficult to gather enough data to conduct the same type of study.
For now, though, a focus on Greenland may be the biggest priority anyway.
"Antarctica is losing ice," Bevis said. "But Greenland is the most unstable ice sheet we have right now."
