Historical Records Provide a Growing Sense of Global Warmth
© 2000 The Washington Post Company
September 8, 2000
By Curt Suplee
Washington Post Staff Writer
An international team of scientists sifting through historical records has found unusual evidence for long-term global warming: Many lakes and rivers in the Northern Hemisphere typically freeze over a week later--and thaw out 10 days sooner--than they did 150 years ago.
Interestingly, that trend appears to have begun at least half a century before the profound buildup of greenhouse gases caused by the burning of fossil fuels by humans.
The team analyzed written accounts, some of them centuries old, including newspaper reports, fur traders' records, ship navigation logs and notes of religious events. From that welter of sources, they extracted the dates of ice formation and breakup in dozens of locations on three continents since 1846.
The result is "a very clear record of the response of aquatic systems to [global] warming," said lead researcher John J. Magnuson of the University of Wisconsin. The findings are reported in today's issue of the journal Science.
In addition, the group concluded, year-to-year variation in freeze and thaw dates appears to have grown over the past three decades, perhaps as a result of powerful El Nino and La Nina episodes and other ocean-atmosphere interactions.
The unconventional historical method provides "an independent, different kind of observational data set, which supports the direct measurements," said Ants Leetmaa, director of the National Oceanic and Atmospheric Administration's Climate Prediction Center, who was not involved in the work.
As part of a decade-long effort to build a global database of river and lake ice information, Magnuson's group "reached out to colleagues around the world," he said, soliciting archival data in a process he described as "network science."
The researchers amassed historical records for 26 sites in the United States, Canada, Siberia, Japan, Finland and Switzerland, and found that the trend in freeze and melt dates was surprisingly consistent. On average, those lakes and rivers are completely covered in ice 8.7 days later in the year than they were in the mid-19th century; and the annual emergence of open water is earlier by an average of 9.8 days.
The records came from a wide variety of sources, including churches on Lake Constance in central Europe. There, the researchers wrote, "the criterion for 'total ice cover' was a walk across the ice of the main basin to transport a Madonna figure between two churches: one in Germany, the other in Switzerland. The figure remained on one side of the lake until the next ice-covered winter, when it was possible to carry it back again."
Twenty-six locations make up a rather small database from which to extrapolate climate changes in an entire hemisphere. But "when you do something that's serendipitous, you take what's there," Magnuson said. Nonetheless, "the wide distribution" of the sites--which range from 36 to 65 degrees latitude--and the "fact that they were more evenly spaced around the globe than similar kinds of observations . . . gave us confidence that the data would represent what was going on at a Northern Hemisphere level."
The magnitude of the date-shifts found in the study, the authors estimate, would require an average temperature increase of about 3 degrees Fahrenheit in this hemisphere over 150 years.
That is larger than, but generally consistent with, the observed 1.2-degree rise in global average surface temperature since the end of the 19th century. Most of the recent warming has taken place in a band of the Northern Hemisphere between 40 and 70 degrees latitude--roughly the area between Philadelphia and central Greenland.
"We know that the regions of Eurasia and North America, where most of these [lake and river sites] are, has indeed warmed according to thermometers at a rate of about double that for the globe--both recently and for the past century," said Kevin Trenberth, head of climate analysis at the National Center for Atmospheric Research, who did not participate in the research.
However, Trenberth noted, it is uncertain whether or to what degree the freeze- and melt-date records accurately reflect global or hemisphere-wide trends. "These data add a little, but not a lot" to understanding such broad changes, he said.
Both Trenberth and Leetmaa emphasized that the warming trend in the study begins even prior to the mid-19th century--long before industrial civilization began pumping enormous amounts of greenhouse gases into the atmosphere. In 1850, world population was about 1.3 billion (approximately one-fifth of the current number) and North America was sparsely populated.
The authors acknowledge that "the few records before 1846 suggest that long-term changes toward later freezing and earlier breakup dates were already occurring, but at slower rates, at sites as far apart as Europe and Japan."
So although some of the temperature trend in the past 150 years is likely the result of human-induced global warming, they write, some "may be related to other [factors] such as changes in solar activity."