Coaxing Nature to Reveal 1,000 Years of Earth’s Climate

By Larry O’Hanlon, Globe Correspondent.
This story appeared in the Boston Globe, May 20, 2003, page C9.

The 20th century may not have been so extraordinarily hot after all, according to a climate study of the last thousand years, which confirms historical accounts of fig trees growing in Germany and early grape harvests in England during medieval times.

The study is part of a fast-emerging field in which scientists combine the data from many natural indicators of past climates to reconstruct what sorts of temperatures and rainfall were experienced over large areas of the globe long before scientific weather records were kept. Such work is providing a much better picture of the past climate, a subject of increasing importance after a century in which the Earth’s average temperature increased by one degree Fahrenheit.

”We felt it was time to pull together a large sample of recent studies from the last five to 10 years and look for patterns of variability and change,” said Willie Soon, an astronomer at the Harvard-Smithsonian Center for Astrophysics.

Soon, along with astronomer Sallie Baliunas and others, integrated more than 240 paleoclimatic studies in order to better understand weather extremes of the past 1,000 years. The team reviewed four decades of research, including work on ancient tree rings, glacier ice cores, sediments from lakes, rivers and seafloors, historical records, coral reefs, stalagmites, fossils, dust, pollen, and many other natural indicators that vary depending on the temperature, precipitation, and other climate factors.

Their work, published in a recent issue of the journal Energy and Environment, suggests that there really was a ”Little Ice Age” from 1300 to 1900 when unusually cold global temperatures froze rivers that today remain ice-free year-round. They also found evidence for a medieval warm period from 800 to 1300 when, historical records show, fig trees grew as
far north as Germany. This 500-year period was warmer than the 20th century, Soon believes.

The finding that the 20th century was not the hottest of the last millennium is controversial in part because it undercuts the argument that burning oil and other fossil fuels is the leading cause of the current temperature increase. However, if the temperature was even warmer before large-scale use of fossil fuels, perhaps the current warm period reflects natural
climate variation. As a result, Soon urges caution in concluding that human action has caused global warming.

”It points to the natural variability of the climate system,” he said in an interview with Canada’s National Post. ”It’s quite likely [some of the warming] is simply due to the dimming and brightening of the sun.”

Other researchers disagree with Soon’s conclusion that the Middle Ages may have been warmer than the 20th century, noting that reconstructing the climate record is a matter of interpretation.

Some even suspect a political motive in releasing the research, funded by NASA, the Air Force Office of Scientific Research, the National Oceanic and Atmospheric Administration, and the American Petroleum Institute. A
Canadian climate scientist told the National Post that the research provides cover for the Bush administration’s opposition to the Kyoto Accord, which attempts to rein in the release of carbon dioxide and other so-called greenhouse
gases that trap heat in the atmosphere.

However, even Soon’s critics agree that studies such as his, which bring together a variety of measures to flesh out climate history, are rapidly coming of age. Over the past eight to 10 years what’s called multi-proxy mapping has become more common, says Malcolm Hughes of the University of Arizona’s Laboratory of Tree-Ring Research. Hughes and his colleagues have just published in the International Journal of Climatology a 1,000-year climate record for New Mexico and Arizona based on two methods of analyzing a set of 19 long-lived trees.

The trick to making sense of the diverse paleoclimate record is to understand that some measures are more precise than others, says Hughes. For instance, rings on trees in Canada may tell you something about summer-to-summer changes in precipitation, while lake sediments, stalagmites, and ice cores may not record every year and give more of a decade-long or 50-year picture.

”It’s a different activity to put all of them together and reconstruct the climate of North America as a whole,” said Kevin Trenberth of the National Center for Atmospheric Research in Boulder, Colo. It takes a lot of collaboration. A tree ring specialist, for instance, knows the strengths and weaknesses of their natural archive, while a climate modeler can piece it all together into simulations that incorporate interactions of the oceans, atmosphere, land surface, ice, vegetation, and water cycles so that it all makes sense.

The danger, says Hughes, is that without the right expertise, climate proxies can be mixed and matched to produce inaccurate pictures of the past. ”You have to compare like with like and you have to be able to calibrate the data,” said Hughes. By calibrate, Hughes means tying tree rings or other climate proxies to real meteorological data, such as thermometer and rainfall records. This requires that the natural archive a scientist is using overlaps a period of time when thermometers and rainfall gauges were in use somewhere nearby — a tall order in many parts of the world.

For example, 2,000-year-old bristlecone pine trees now living on a mountaintop in Nevada can have a core extracted (without harm to the tree) and the most recent annual rings compared to meteorological records of the area dating back about a century. Once a century of tree rings can be calibrated to local temperature and rainfall records, and the details of how the rings change with the changing conditions, the process can be reversed and the rest of the tree ring record can be deciphered to reveal what sort of rainfall and temperature patterns were present for 2,000 years.

The process quickly gets complicated when proxies from many areas are brought together to map out ancient climate trends for a region. Because different proxies yield varying pictures of changes through time, some paleoclimate maps can look vague and generalized. It’s also a matter of having enough computer power to handle not only the data, but to incorporate it with the oceanic, atmospheric, and other models, says Hughes.

But the results will be worth it. Once the millennial records of changes in oceans, land, and air are sorted out, they can be used to fine-tune computer climate models that take all the same parts of the puzzle into account, says Trenberth. And once those models can be made to reproduce the past climate changes, they can begin forecasting what might happen next. No one knows how accurate such models will be, but it’s certain they will need a lot of data and computer power to take into account all the variables that influence the Earth’s climate.

Meanwhile, the natural climate archive is disappearing faster than scientists can gather data. ”Ice is melting, corals are being destroyed, trees are being cut down,” said Trenberth. ”Collecting needs to speed up.”

Partner & Fellow Blogs