DURHAM - For more than a decade, the federal government has spent millions of dollars pumping elevated levels of carbon dioxide into small groups of trees to test how forests will respond to global warming in the next 50 years.

Some scientists believe they are on the cusp of receiving key results from the time-consuming experiments.

But the U.S. Department of Energy, which is funding the project, has told the scientists to chop down the trees, collect the data and move on to new research. That plan has upset some researchers who have spent years trying to understand how forests might help stave off global warming and who want to keep the project going for at least a couple of more years.

"There has been an investment in these experiments, and it's a shame we are going to walk away from that investment," said William Chameides, an atmospheric scientist at Duke University, where one of the experimental forests is located.

Ronald Neilson, a U.S. Forest Service bio-climatologist in Corvallis, Ore., said the experiments should continue because they still have potential to answer key questions about how rainfall and fertility affect how much carbon a forest will store longterm -- essential to understanding how forests might soften the blow of climate change.

But the Energy Department, following the advice of a specially convened panel of experts, thinks that chopping down the trees and digging up the soil will allow the first real measurements of how much carbon the leaves, branches, trunks and roots have been storing, said J. Michael Kuperberg, an agency program manager.

Ending the experiments also will allow the funding to be devoted to new research that will look at the effects of higher temperatures, changes in rainfall, and variations in soil fertility, Kuperberg said.

The research program, Free Air CO2 Enrichment (FACE), consists of rings of tall white plastic pipes with holes along their length that emit once-liquefied carbon dioxide in carefully metered doses. The carbon dioxide levels around the trees are about 50 percent higher than current levels -- the amount expected 40 to 50 years from now.

Carbon dioxide enrichment of the loblolly pines, planted in 1983 several miles from Duke's campus, began in 1994. There are also experiments in Tennessee and Wisconsin.

Results so far indicate that elevated levels of carbon dioxide make forests grow more quickly, said Ram Oren, professor of ecology at Duke University's Nicholas School of the Environment and Earth Sciences and principal investigator on the experiments there.

But unless forests are on fertile ground -- hard to come by because of development -- growth will be in leaves, needles and fine roots, which die off and decompose in a year or two, releasing the carbon dioxide back into the atmosphere, Oren said.

"To stop an experiment that cost $55 million $10 million before it reaches its real conclusion makes no sense to me," Oren said.