A signature piece of evidence for global warming -- a compilation of data showing that a sharp rise in temperatures made the late 20th century the warmest period in 1,000 years -- is probably true, a national panel of scientific specialists concluded yesterday.
A graph of the data has become an icon of global warming and is often referred to as ``the hockey stick" because of its shape: A shaft that shows a long period of relatively little change in Northern Hemisphere surface temperatures, and then a spike upward during the last 100 years or so that resembles the blade.
Since the first version of it was published in a scientific journal in 1998, environmentalists have seized on the graph as powerful evidence of human-induced climate change, while some critics have called it alarmist, questioning its methodology and the accuracy of its temperature data.
Last year, the dispute catapulted into the national political arena after Joe Barton, a Texas Republican who is chairman of the House Energy Committee, asked the three authors of the 1998 study -- including a University of Massachusetts professor -- for a detailed accounting of their government and private funding, data, and methods. A range of scientists and other legislators blasted the request as an intimidation tactic, contending that other researchers would be reluctant to embark on such studies if they knew they would be under such scrutiny by members of Congress.
The National Academy of Sciences, which advises Congress and the government, was then asked to conduct an independent review by the chairman of the House Science Committee, Republican Sherwood L. Boehlert of New York.
``Our conclusion is that this recent period of warming is likely the warmest in the last millennium," said John M. Wallace, one of the 12 panel members and a professor of atmospheric sciences at the University of Washington.
During a Washington press conference yesterday, other members of the panel said that they had a high level of confidence -- 90 percent to 95 percent -- that the planet is in its warmest period in 400 years and that the odds are ``2 to 1" that this is probably the warmest period stretching back 1,000 years, as the original study concluded.
The panelists had far less confidence in the reliability of temperature data from longer than 1,000 years ago; the initial study did not look that far back. And they said that because it was difficult to reconstruct exact temperatures for specific years, they were unable to support the study's contention that the 1990s were the warmest decade, and 1998 the warmest year, at the time of the study , in the past 1,000 years.
Still, Wallace said, the general trend holds true: ``This doesn't change the scientific landscape in terms of the greenhouse warming debate," he said.
Scientists widely believe that power plant and car emissions of carbon dioxide that become trapped in the atmosphere are the primary cause of the earth's temperature rise of about 1.4 degrees Fahrenheit in the past century. The 1998 study supports that view -- the rise in temperature coincided with the Industrial Revolution -- but scientists have amassed other long-term evidence, including the retreat of glaciers and the melting of polar ice caps.
Because reliable temperature records stretch back only about 150 years, scientists must infer past temperature and climate data from natural archives, such as tree-growth rings, corals, ice cores, and cave deposits. For example, trees at high latitudes and altitudes grow faster during warmer periods, so scientists can determine how warm it was from the size of growth rings. The thickness of sediment layers at the bottom of certain lakes also provides temperature clues because the increase in snow melt in warmer periods washes more sediment into lakes. Scientists also examine other records, such as historical accounts of wine harvests or paintings of glaciers in northern regions.
Scientists had been gathering data from these indirect methods for years, but they had not been put together until Michael Mann of Pennsylvania State University, Raymond Bradley of the University of Massachusetts at Amherst, and Malcolm Hughes of the University of Arizona created a statistical model that enabled them to construct a 1,000-year temperature record and plotted the results on an easy-to-read graph. A version of the graph going back 600 years was published in their 1998 study in Nature. An expanded one stretching back 1,000 years was part of a landmark 2001 international report on global warming that concluded humans were driving the rise in temperatures. The graph, the hockey stick, soon became the most recognized statistical visual of global warming, reproduced in media accounts the world over and more recently in former vice president Al Gore's movie ``An Inconvenient Truth."
It was also seized upon by critics. An economist and a minerals consultant in Canada uncovered inconsistencies in the model used to create the graph and questioned the use of certain data they said biased the results. Critics also said the methodology might have underestimated past periods of warming, especially a warm period that may have started about 1,000 years ago. If that is true, it would mean that the earth may have had a wider natural climate swing in the past millennium than scientists had believed -- possibly indicating the current warming trend is not such an anomaly.
Yesterday, panel members said there was isolated evidence of a warmer earth about 1,000 years ago, but it appeared regional. They suggested that it did not reflect the overall global temperature at the time. They also said that while there were certain biases in the report, they got roughly the same results as the initial study when they used different methodologies to put together the data. The report said that Mann's study had been supported by ``an array of evidence" in other studies and that his and his collaborators' findings were not even a primary piece of evidence of human-induced global warming.
Panel members said that while the earlier study was not perfect, Mann and his colleagues should be praised as the first to devise such detailed global temperature measurements and that its benefits far outweighed any detriments. However, in a nod to the difficulty some critics had in getting data from the original researchers to replicate the initial study, they urged climate researchers to more readily share their data. The panel also called for more research to improve the reliability of temperature estimates from before 1600.
``The hockey stick is alive and well," said Bradley, director of the University of Massachusetts at Amherst Climate System Research Center. He said the original research paper was never meant to be taken as absolute and even included the words ``uncertainties and limitations" in its title. ``Nothing changes the general shape."
The National Academy of Sciences study did not mollify critics, however, who said that great uncertainty still existed about temperatures prior to 1600. Those critics agree that the scientific data show a distinct warming from about 400 years ago until now, but ``substantially larger uncertainties" exist looking further back, said Ross McKitrick, an economist at the University of Guelph in Canada who has been one of the primary critics of the study.
A spokesman for Barton said a separate group of statisticians recruited by his energy committee was still examining the 1998 study. Barton spokesman Larry Neal said `` all we want to know is whether the numbers add up."
Beth Daley can be reached by e-mail at email@example.com.