FAIRBANKS — A new digital atlas chronicling sea-ice concentrations in the Beaufort, Chukchi and Bering seas from 1850 to 2012 has been launched.

The interactive, Web-based graphic will allow users to view the extent of sea ice as much as 300 miles off Alaska’s coast, dating back to 1850. Users can search based on date or location, and there are animations of monthly ice change, ice-coverage graphs and a glossary of sea ice and scientific terms.

The atlas is currently only searchable back to 1953, but the remaining portion should be live next week.

The atlas was compiled by the Alaska Center for Climate Assessment and Policy (ACCAP) at the University of Alaska Fairbanks and the Scenarios Network for Alaska and Arctic Planning (SNAP) and combines more than 10 historical and current databases.

Anchorage-based Alaska Ocean Observing System funded the $200,000 project.

John Walsh, chief scientist for UAF’s International Arctic Research Center, and Sarah Trainor, director of ACCAP, led the two-year effort. Walsh said researchers wanted a better, more reliable tool to understand the sea-ice cycle. 

“If you only have 30 years of data, it’s really hard to say there is a 20-year cycle,” Walsh said.

The first 100 years of information rely mostly on log books and charts from whaling ships, provided by the New Bedford Whaling Museum, and ice surveys by the Danish Meteorological Institute. The National Snow and Ice Data Center scanned almost 7,000 charts made between 1953 and 1986, donated by the estate of Bill Dehn, who did exploratory work and surveys for the Navy and oil companies. Satellite coverage began in 1979 and provides accurate weekly readings.

“There are a lot of uncertainties,” Walsh said of using old records. Whaling ships would simply record their location and note if ice was visible or not. Reports could vary greatly depending on time of month, and it created challenges coming up with monthly estimates in a meaningful way.

The final atlas will include two versions. The first will only use physically recorded information, the second will be labeled “no data,” so users know it’s extrapolated.

“It’s an important distinction,” Walsh said. “We don’t want people taking what we’re calling these estimated data areas as absolute truths.”

Researchers used an analogue approach, where they supplement fragmented data with more complete information from similar conditions. If there was a month with incomplete data the team would look for the most comparable ice coverage in the past 50 years, find the four or five best fits, and fill in the missing coverage with the analogous, or similar,


One of the first things Walsh wanted to investigate was any possible precedent for the extreme lack of summer ice during the past five or six years.

“There were occasional one-year retreats, in 1958 and 1968, where the ice edge was 300 or 400 miles back from Barrow,” Walsh said. “There is no precedent … for ice retreat as extreme as we’ve had.”

The atlas can be seen at seaiceatlas.snap.uaf.edu.

A webinar hosted by ACCAP is scheduled at 10 a.m. Tuesday. It will include samples of input information and discuss the sources and procedures used. For webinar information, visit accap.uaf.edu/node/1048 or call Tina Buxbaum at 474-7812.