Scientists have theorized that the loss of Arctic sea ice over the last three decades is part of a recent, dramatic change in global climate. Now they have proof from an unorthodox collection of sources.

Researchers with ARCdoc, a project based at the University of Sunderland in England, found that annual sea ice between Canada and Greenland blanketed much more territory up to 250 years ago—the result of a colder, wetter, stormier climate. The evidence came from nearly 150 handwritten logbooks of Royal Navy expeditions and Hudson’s Bay Company transports in the North Atlantic between 1750 and 1850. Another 60 volumes from Arctic whalers, who hunted their prey among permanent ice floes, offered rare eyewitness accounts of the far north. While the naval and Hudson’s Bay records had been examined before, “These whaling logbooks had never been used for scientific studies,” says Dennis Wheeler, ARCdoc’s director.

The three-year endeavor to render the musty annals into numerical data was undertaken in two phases: First, translate the books’ observations of ice forms into a modern glossary. Second, reconstruct a vital piece of the North Atlantic climate prior to anthropogenic warming.

Crunching the notes on precipitation, ocean activity and wind required a crash course in venerable maritime argot. ARCdoc research assistant Matthew Ayre focused his efforts on a handful of books by William Scoresby, Jr., one of Britain’s most ambitious whalers and an early polar scientist. Ayre prepared a sea ice “dictionary” by translating the archaic lingo into present-day terminology and percentages. Then, he validated its accuracy with regular observations from the deck of an icebreaker in the Arctic Ocean.

Armed with this polar decoder ring, the ARCdoc team painstakingly extracted meteorological memos from the logbooks, updating the antiquated language to the scientific standard. Finally, they organized monthly tallies of weather events and seasonal ice characteristics and mapped the data across Baffin Bay and Davis Strait—an area known as an incubator for climate trends.

The sequence of data points, called a time series, revealed region-wide decadal changes in wind circulation, storm frequency and intensity, and precipitation. And, around 1800, a major shift in storm strength brought more blizzards and ice cover for the next 40 years. Baffin Bay and Davis Strait saw greater expanses of sea ice and bigger swings in weather patterns than scientists had assumed.

After comparing their model with current satellite images, the ARCdoc team confirmed a hefty decrease in the ice’s reach over two centuries, the first proof of the frostier past theorized by climatologists. Moreover, ARCdoc’s results show a section of the Arctic before the greenhouse effect, allowing scientists to gauge current and future conditions against the “natural” (preindustrial) environment.

The records focus on a single region during a limited time period but demonstrate the range of variability within the climate system, says Walt Meier, a research scientist at NASA Goddard Space Flight Center. ARCdocs’s findings could be linked to other data sets for a bigger picture of atmospheric phenomena beyond the North Atlantic. “It’s really valuable for creating context for the longer-term variations,” he says.

Mark Serreze, director of the National Snow and Ice Data Center, called the study a great example of data-mining—in this case, plumbing historical narratives to discover patterns among raw data. “This project affirms that what we’ve seen in the satellite data is highly unusual, and the findings could be a useful tool for showing the public that things today really are different,” he says.

But some questions remain: Did volcanic activity, solar radiation, or oceanic circulation contribute to ice loss? Wheeler says clues might be gleaned from logbooks in American, Dutch or Danish collections that have yet to be studied. The answers could give scientists the means to build precise models of our planet’s most climatically sensitive place. If we have accurate knowledge about the Arctic’s past, we can make better decisions for its future.