Arctic sea ice data from satellites only goes back to 1979, so evidence of recent thinning has tended to lack historical context. But Science Daily recently reported on a new study using satellite data, ship logs and computer modeling that reconstructed Arctic sea ice thickness and volume back to 1901. The result was an inverted hockey stick which implies the sea ice volume decline in recent decades is dramatic and unprecedented. Global Warming! But as Ron Clutz points out, ship logs go back to the early 1800s, and a previous comparison making use of the pre-1900 evidence found ice conditions during the 1800s were remarkably similar to the present.
As always with climate, modeling Arctic sea ice is not easy. Ice is affected by air temperatures, which are hard enough to measure, but also by winds, ocean currents, snowfall and so forth. Some years ago experts at the Institute of Ocean Sciences in Victoria, BC, argued that Arctic sea ice hadn't thinned as much as was commonly believed. The ice was mostly still there, it had just piled up in different parts of the Arctic basin due to changing wind patterns. Their conclusion was based on modeling work. Another modeling study a few years later concurred that wind was partly to blame but spring air temperatures and increased ice flow out of the Fram Strait between Greenland and Iceland also played a role.
The new study likewise is an exercise in modeling, built in this case around a set of ice thickness measures going back to the late 1940s, mostly from the Canadian Arctic. What's remarkable about the data (click on the link: red is from the model, blue from the Canadian observations) is how stable the average is over time. Sea ice thickness declined after 1990, but prior to that it had been rising since the 1950s. (The dots in the 1920s are from the expedition of the Maud from 1922 to 1924.)
If ice thickness hasn't changed, why does the ice volume look like a declining hockey stick? Because the ice surface area has gone down. Why has it done so? Well, now we're back to trying to figure out how much is due to temperatures, how much to wind, currents, etc.
This is where the article cited by Clutz becomes important. By overlaying past sailing routes on modern ice extent diagrams, they concluded:
The extent of summer sea ice during the 19th century, insofar as it is shown in patterns of navigability inferred from ship tracks, the direct observations of explorers, and a number of native accounts, is remarkably similar to present ice climatology
So yes, the extent of sea ice in recent years has gone down. But only to levels familiar to sailors in the 1800s, a period nobody associates with global warming of any sort, let alone man-made or catastrophic. Thus those authors dispel the idea that the Arctic weather was always the same until the late 20th century. Rather, in some years early explorers encountered unusual warmth and almost made it through the Northwest Passage, while in others the sailing season was short and winter came early. But overall, 19th century sailors would consider today's Arctic routes very similar to what they had encountered. The researchers also note that the few measurements of average ice thickness we have from 19th century voyages are close to modern measurements from the same locations.
In sum, while scientists are still trying to figure out what explains changes in Arctic ice cover, and how much increased air temperatures might change the summertime ice extent in future decades, there is still a lot to learn about how much it has varied in the past and what the natural drivers of variability are.