A new NASA study has revealed a new mode of ice loss in Greenland that may increase the potential for sustained loss of mass as the climate continues to warm up, with implications for the future rate of sea level rise.

During Greenland's hottest summers on record, 2010 and 2012, the ice in Rink Glacier on the island's west coast did not just melt faster than usual, it slid through the glacier's interior in a gigantic wave, like a warmed freezer pop sliding out of its plastic casing, showed the findings, published online in the journal Geophysical Research Letters.

The wave persisted for four months, with ice from upstream continuing to move down to replace the missing mass for at least four more months. This long pulse of mass loss is called a solitary wave.

"We know for sure that the triggering mechanism was the surface melting of snow and ice, but we do not fully understand the complex array of processes that generate solitary waves," said Surendra Adhikari from NASA's Jet Propulsion Laboratory in Pasadena, California.

The study by the JPL scientists was the first to precisely track a glacier's loss of mass from melting ice using the horizontal motion of a GPS sensor. 

They used data from a single sensor in the Greenland GPS Network (GNET), sited on bedrock next to Rink Glacier. 

Rink is one of Greenland's major outlets to the ocean, draining about 11 billion tonnes (gigatons) of ice per year in the early 2000s.

In the intensely hot summer of 2012, however, it lost an additional 6.7 gigatons of mass in the form of a solitary wave. 

Previously observed melting processes cannot explain that much mass loss.

The wave moved through the flowing glacier during the months of June through September at a speed of about four kilometres a month for the first three months, increasing to 12 kilometres during September. 

The amount of mass in motion was 1.7 gigatons, plus or minus about half a gigaton, per month. Rink Glacier typically flows at a speed of a few kilometres a year.

The researchers saw the same wave pattern in the GPS data for 2010, the second hottest summer on record in Greenland. 

Although they did not quantify the exact size and speed of the 2010 wave, the patterns of motion in the GPS data indicate that it must have been smaller than the 2012 wave but similar in speed.

"What makes our work exciting is that we are essentially identifying a new, robust observational technique to monitor ice flow processes on seasonal or shorter time scales," Adhikari, who led the study, said.