What’s in a number? Arctic sea ice and record lows

Arctic sea ice extent for September, 2011 was the second-lowest in the satellite record. However, other data sources showed that ice extent perhaps hit a new record low for several days in September. Credit: NSIDC/NASA Earth Observatory
Did Arctic sea ice reach an all-time record low this year—or not? Scientists at University of Bremen in Germany thought it was a new record, while data from NSIDC showed the sea ice at its second-lowest level. The two groups were examining data from two different satellite sensors. Why did the data differ—and if data can vary, does a record low really matter?

Different data, different story?

NSIDC scientists report ice extent, a two-dimensional measure that shows how much of ocean surface is covered with sea ice. NSIDC uses data from Department of Defense satellites, which go back to 1979. Over the last 30 years, this data set has shown a major decline in ice extent, which is more pronounced in summer months.  Around that downward trend line, ice extent oscillates up and down, influenced by weather, winds, and currents that change from year to year. Other measurements, from different satellites, airplanes, and fieldwork, show the same long-term decline as the NSIDC data, but the exact numbers can vary.

This fall, analysis by scientists at the University of Bremen, using data from the Advanced Microwave Scanning Radiometer-Earth Observing System  (AMSR-E) sensor, on board the NASA Aqua satellite, showed sea ice reaching a new record low for a few days in September. NSIDC scientist Walt Meier explained that the difference came from the higher resolution of the newer AMSR-E sensor: because the sensor provides more detail, it shows areas of open water that are classified as ice-covered by the lower-resolution sensor that NSIDC uses. But the AMSR-E data only go back to 2003, making it harder to compare those data to the long-term record. Meier said, “In the end, these are minor differences. The bottom line is that all sea ice data sets are telling the same story: the Arctic Ocean is losing sea ice.”

Breaking records

If different satellite sensors can disagree on whether a record was reached, what does a record low really mean? “Whether this year was lowest or second-lowest isn’t important in the big picture of climate change,” said Meier. What does matter is the long-term trend. He said, “The long-term trend is becoming more and more clear: sea ice has declined by 12 percent per decade since 1979, and the last five years have still been the five lowest in the satellite record.” So while new records can attract attention to declining sea ice extent, researchers say that focusing only on these milestones can be misleading. “Records are always an interesting marker,” said Meier, “They're worth noting, but they only really have meaning in the context of the long-term trend.”

Like record-breaking heat waves or snowstorms, sea ice extent is influenced by weather conditions in the Arctic that vary from year to year. So while researchers expect the decline to continue, nobody expects to see new records every single year. Stefan Rahmstorf from the Potsdam Institute for Climate Impact Research, pointed out in a recent article that the record or near-record ice extent this year reinforces the idea that the low ice extent in 2007 was not just an anomaly. That downward trend means big changes for the Arctic Ocean—from increased shipping to changing wildlife habitats. Perhaps the biggest wildcard is the potential for even greater warming as the additional open water absorbs more heat from the sun. For more information, read Arctic Sea Ice 101.