The scent of Black Spruce filled the crisp autumn air as Scott Goetz made his way to the array of light detectors spread beneath the canopy of the Alaskan forest. The sun sat low on the horizon, casting long shadows on the forest floor. The long night of winter was quickly approaching, and the detectors would no longer be needed for the year. Though he was there to study the forest, how it absorbed light and carbon and turned them into materials like leaves and bark, it was the wilderness of the place that initially attracted Goetz. He had grown attached to these Northern forests as a graduate student canoeing through the interlinked lakes of Minnesota’s Boundary Waters Wilderness Area while collecting data for a NASA research project. Now an established ecologist, Goetz had spent at least a decade studying and exploring the boreal forests of North America. He knew the forest well.

At the close of the growing season of 2004, Goetz was seeing changes in the forest. The blackened skeletons of trees extended into the distance after fires consumed a record six million acres. Up close, the gold and green leaves of the rustling aspen trees were marbled with white squiggly lines where leaf miners had eaten through them. While both fires and insects were a natural part of the forest’s lifecycle, Goetz hadn’t seen either have such a wide impact on the forest before. But the most disturbing and unexpected change, he had observed months earlier back in his office at the Woods Hole Research Center in Woods Hole, Massachusetts.

Spruce trees surround light detectors on the floor of the boreal forest in Alaska. Scientists are traveling to sites in the remote North to study the effects of climate change on ecosystems around the world. Scott Goetz uses light detectors to measure the sunlight absorbed by the trees. The scientists combine these measurements with satellite data to gauge the health of the forest. (Photo copyright Daniel Steinberg, Woods Hole Research Center.)

In addition to his research in the forest, Goetz was using satellite data to study how the spruce-rich forests of northern Canada and Alaska recover after large fires. The burned forest was re-growing as he expected, but the unburned forest was behaving strangely. Since the 1990s, scientists have known that increasing global temperatures have lengthened the growing season in the Arctic. With carbon dioxide, one of the key ingredients in photosynthesis, also on the rise, the forest should have been thriving. But it wasn’t. The forest was getting browner, not greener.

On the other side of the United States, Alon Angert, a scientist at the University of California, Berkeley, noticed a strange trend in the forest, too. Angert was tracking carbon dioxide in the atmosphere over the Arctic from 1985 to 1994 when he saw that trees weren’t soaking up as much of the gas at the end of the period as at the beginning. It was as if the whole forest had slowed its breathing during that single decade.

“Something big is happening in the high latitudes,” says Rama Nemani, a research scientist at NASA Ames Research Center, in response to the papers that Goetz and Angert published within weeks of each other. Nemani was on the research team that initially noticed that the Arctic was beginning to green in response to global warming in the 1990s. Despite the previous discovery, Nemani wasn’t surprised that Northern forests now seem to have slowed their growth. After all, the same theories that predicted that global warming would increase forest growth in the Arctic, theories that Nemani helped prove, also predicted that the forests would eventually reach the limits of the water supply and go into decline. “We knew something like this would happen,” Nemani says. “We didn’t expect that it was going to happen so quickly.”

In early autumn, the brilliant golden leaves of an Aspen forest cover hillsides in Alaska, while conifers dominate at lower elevations. Insect infestations and other signs of ill-health are starting to appear in the forest, signs of the changing climate. (Photo copyright Robert Ott.)

The boreal forests (also called taiga) encircle the Arctic, occupying the northern expanses of Asia, Europe, and North America. In this map, the boreal forests are dark green areas, tundra and barren land are tan, while crops and grasslands are yellow. In total, the boreal forest covers 16.6 million square kilometers (6.41 million square miles.) (Map by Robert Simmon, based on data provided by BU Land Cover and Land Cover Dynamics.)

What is happening to the forests of northern Alaska, Canada, Europe, and Siberia? Why have they slowed their growth when everyone thought they should be expanding for several more decades? Is the trend that Goetz and Angert independently observed a fluke, a temporary downturn in the health of the forests, or is it something more? Is it a sign that global warming is changing Northern forests more quickly than anyone thought possible?

The answers to these questions impact more than the sparsely populated forests of the North. In most predictions of global warming, the forests of the far North will expand into the treeless tundra and grow more quickly as warmer temperatures improve growing conditions. A fast-growing forest would soak up more carbon dioxide, dampening the effects of warming. But if the forests aren’t growing as predicted, carbon dioxide could build in the atmosphere, driving global temperatures up even more. Warming temperatures could bake soil and plants, causing them to release more carbon dioxide than they absorb. The switch from carbon sink to source is a possibility that scientists anticipated for the next century, but are Goetz and Angert’s observations a sign that the forests are already approaching this threshold?

 Satellites Reveal a Browning Forest

Satellites Reveal a Browning Forest

These were not the questions that Goetz set out to answer when he began his analysis of satellite imagery of the Northern forests of North America. He had wanted to know how the forests use carbon dioxide as they recover from fire so he could determine the impact of fire on the carbon cycle. As Goetz expected, the satellite data showed that the newly burned forest was “greening up” as it recovered to pre-burn conditions. But in the surrounding unburned forest, growth was slowing down, and that surprised Goetz. “Earlier work suggested greening in the Northern Hemisphere and an increase in the growing-season length,” says Goetz.

He decided to take a closer look at the unburned forest, enlisting the help of Woods Hole colleagues Andrew Bunn, an ecologist with experience tracking change in the forests, and Richard Houghton, a senior scientist and carbon expert who had spent years studying the global carbon budget.

The group studied data collected by the Advanced Very High Resolution Radiometer (AVHRR) series of satellite sensors, which have been in orbit since 1981. When sunlight hits the surface of the Earth, some of it bounces back into space, and this is what the satellites measure. “Plants reflect light depending on how fast or how much they are growing,” says Bunn. “The different light that they reflect depends on how green they are and how much photosynthesis is happening. By measuring that greenness, we can infer how much production is happening.” Production, he explains, is how much carbon plants take out of the atmosphere and turn into organic material like leaves and stems through photosynthesis.

To measure forest “greenness” with satellite data, scientists often create a low-to-high vegetation scale, or index. The Global Inventory Mapping and Monitoring (GIMMS) group at NASA Goddard Space Flight Center has produced such vegetation index maps for every 15-day period that the AVHRR sensors have been in space. To track the change in Northern forests’ greenness, Goetz and research assistant Greg Fiske gathered all of the 15-day vegetation index maps that the GIMMS group had produced for 1981 to 2003. “We stacked all the 15-day periods on top of each other and looked at how each point in space was changing over time,” explains Bunn. The trend that emerged revealed a browning forest. “I was surprised,” says Goetz. “We looked at it a couple of times. In fact, we reprocessed the data twice and then a third time to be sure.” It was no mistake; the forest was browning, but only in interior Alaska and Canada. “When we started to look at a much larger area, we saw that the tundra areas continued to green, where the forest areas were in decline.”

Dan Steinberg, a member of Goetz’s team, stands in a boreal burn scar. Scott Goetz and his colleagues first noticed changes in the health of the boreal forest while studying burn scars. While burned areas regrew as scientists expected, growth slowed in the surrounding unburned land. (Photograph copyright Daniel Steinberg, Woods Hole Research Center.)

But why were the forests in decline? Scientists have always thought that plant growth in the boreal forests was limited by temperature. Arctic summer provides a brief period in which plants can thrive before the cold of winter ends the growing season. If temperatures had warmed, extending the growing season, then plants should have been able to grow more. But Goetz and his colleagues suspected that warmer temperatures had also dried the forest.

“Most people wouldn’t think of these boreal forests as being drought stressed,” says Goetz. “There is often a lot of surface moisture in wet areas, but if the air is very dry, conifers tend to be pretty strongly affected. Photosynthetic rates are reduced when the air is dry, particularly in these high-latitude forests adapted to cooler conditions.” Although drought would cause the same kind of browning that Goetz and his colleagues had observed, they weren’t certain that dryness was the only thing affecting the trees. It was possible that other factors, such as nutrient stress or insect damage, could be to blame. Strong evidence that drought was really to blame would come from a second source.

From 1982 through 2003, the photosynthetic activity (an indicator of plant growth) of the boreal forest in North America decreased. At the same time, the photosynthetic activity of tundra along the Arctic coast increased. This map shows areas of decreased photosynthetic activity in brown, and increased photosynthetic activity in green. Despite warmer temperatures and a longer growing season, the growth of the forest was slowing down. (Map adapted from Goetz et al. 2005.)

Warm Summers Slow Carbon Uptake

On the other side of the country, Alon Angert was puzzling over the same question as Goetz: what was happening to Northern forests? Angert, a researcher at the University of California, Berkeley, had been using remote sensing and carbon dioxide measurements to see how the biosphere—life on Earth ranging from plants and animals to soil microbes—was reacting to climate change. “We know that about quarter of the CO₂ that humans emit into the atmosphere is taken up by the land biosphere, but we want to know if the biosphere will keep taking up CO₂,” he explains.

Angert was particularly interested in finding out how the change in season affected the way plants use carbon dioxide. To get an idea of what carbon dioxide levels were on a global scale, he used carbon dioxide levels taken from all ground stations and averaged by latitude, focusing on latitudes above 20 degrees North where seasonality impacts plants the most. Next, he removed the long-term “background” increase in carbon dioxide that has been occurring since the Industrial Revolution to isolate the seasonal increase and decrease caused by the biosphere “breathing.” Then he divided the data into two periods: spring when plants are coming to life and summer when plants are taking up the most carbon dioxide.

“What we found is that the spring uptake was very highly correlated with temperature,” Angert says. The warmer the spring temperatures were, the more carbon dioxide the forest absorbed, just as theories had predicted. The trend for the summer was a different story. From 1985-1991, the forest did continue to absorb more and more carbon dioxide as it greened throughout the summer. But then, in 1991, Mount Pinatubo erupted, sending a cloud of sulfate aerosols into the upper atmosphere. The eruption cooled global temperatures for two years, disrupting the trend. When temperatures resumed climbing in 1994, a new trend began to emerge. The forest continued to soak up more carbon dioxide with warmer temperatures in the spring, but in the summer, the trees stopped using as much carbon dioxide. “The biosphere didn’t keep up,” says Angert.

“I was surprised,” Angert remarks. Like Goetz and his colleagues, Angert checked his data again. Still seeing the shifting trend, he decided to look at vegetation index data to confirm his observations. Had the forests browned while they slowed their intake of carbon dioxide?

After getting the same satellite-based vegetation data from NASA that Goetz and his colleagues used, Angert divided the data into spring and summer periods to match his carbon dioxide observations. He then averaged the vegetation index values by latitude and saw that they echoed the carbon dioxide measurements. In the spring and the summer of the first period, the forests got greener with warmer temperatures. But in the summer of the second period, when he had seen a decrease in the amount of carbon dioxide the biosphere absorbed, the forest got browner.

Before the eruption of Mt. Pinatubo (yellow block), carbon uptake in Northern forests (blue lines) increased as temperatures increased (green lines) in spring (top graph) and during the entire growing season (March-August, bottom graph). When the cooling period caused by Pinatubo’s emissions ended, global temperatures began to climb again. Increases in temperature continued to boost carbon uptake in the spring, but over the entire growing season, warmer summers caused carbon uptake to decline. The light-colored lines show year-to-year changes, while the darker lines show the overall trend. (Figures adapted from Angert et al. by Robert Simmon.)

“When you see the same thing happen in independent data sets, it makes you think that what you are seeing is real,” Angert observes. As an extra check, he put the satellite vegetation “greenness” measurements into a model to estimate how much carbon dioxide was taken from the atmosphere. The model predictions matched what he had seen in the observations of carbon dioxide.

Why would the forest head into this decline? Like Goetz, Angert suspected drought. The forest was doing fine in the spring, but by summer had run out of water, and that was causing the browning trend, he theorized. He matched the vegetation index data to the Palmer Drought Index, a measure of drought that compares the amount of water that is available to the amount of water that plants need at certain temperatures. Again he found a match. The forest productivity had decreased when the drought index indicated water shortages.

Arctic temperatures warmed about 0.3°C over the past 25 years. During the initial decade of warming, boreal forests responded with vigorous growth. From 1994 to 2002, however, growth in many places slowed as temperatures climbed and the forest dried out. The maps above show trends in greenness, or growth (top pair), temperature (middle), and drought (bottom) from 1982-1991 and 1994-2002. In the earlier decade, increased growth (green) was linked to warmer temperatures (red) across most of the North. In the second decade, most of the area experienced poorer growth (brown) as temperatures rose. A cooling trend (blue) limited growth in Alaska during the latter decade, but the long-term trend in the area is a warming one. A Palmer drought index map shows areas that experienced drier (orange) or wetter (purple) summers. (Maps adapted from Angert et al. by Robert Simmon.)

Ground Observations Support Satellite View

As much as Angert and Goetz were surprised by their findings, ecologists working in the field in Alaska had been expecting just such an analysis to come out. Glenn Juday is a professor of forest ecology and director of the Tree Ring Laboratory at the University of Alaska Fairbanks, and he was the lead author for the forest section of the Arctic Climate Impact Assessment report, a comprehensive summary of scientific knowledge about the Arctic and climate change published in 2004. He has spent much of his career studying how trees in the boreal forests respond to warming temperatures by mapping the correlation between tree rings and temperatures. When asked if Goetz and Angert’s observations matched what he saw in the field, Juday responded with an unequivocal “absolutely.” He has found, for example, that Northern White Spruce grow more slowly after July temperatures reach 16.5 degrees Celsius or warmer. Along the cold, high-elevation tree line, as many as 40 percent of the trees grow less when temperatures are warm. “These are the trees that are supposed to do well as a result of climate warming,” Juday points out. “Well, the trees don’t know that.”

Juday had sampled thousands of trees in a wide range of locations to understand how each type of tree in each type of location would respond to climate change. He thought he had a good understanding of individual species, but not how the entire forest ecosystem was reacting to warming temperatures and a lengthening growing season. “The big problem we were facing was, if you’re drilling cores in trees and taking them back to the lab, and then measuring them, it’s a daunting task to try to cover enough terrain to get an answer for the overall boreal forest of North America,” says Juday. “You can do all the plot sampling feasible in a career, and you still can’t get enough coverage. At some point, you have to have a synoptic view of what’s happening over the planet. The only planet-sampling tools we have are satellite remote sensors.” Goetz and Angert’s analyses of the satellite data told him that a large part of the forest was responding negatively to warming temperatures.

The Boreal Forests Face Climate Change

But what does a drying, browning forest mean? What impact could a decline in the health of the boreal forests have on the rest of the planet? One of the reasons scientists are so concerned about these Northern forests is the large role they have in the global carbon cycle. When carbon dioxide is released from fossil fuel burning or other sources, about half of it stays in the atmosphere where it acts as a greenhouse gas heating the surface of the Earth, and half is absorbed by the land or the oceans. Though scientists aren't sure exactly where all carbon is stored on land, the boreal forests, most of which are intact, are a likely candidate. A little over 30 percent of the world's forests are in the far North, with boreal forests covering 17 percent of the Earth's land surface area. As such, the Northern forests are a big storage area for carbon, says Houghton, the ecologist and carbon expert working with Goetz. If the forests stop soaking up carbon dioxide or slow their productivity, then more carbon dioxide remains in the atmosphere to contribute to global warming.

“Nature’s been helping by storing carbon on land. It’s possible that nature won’t always help us, and we’re starting to have the first suggestions of that,” says Houghton. “The Northern terrestrial carbon sink is starting to waiver. We don’t know if that will continue or get worse, but this seems to be the beginning.” Houghton acknowledges, with some skepticism, “It might be that the last five or ten years were peculiar.” It’s possible, in other words, that the warming and drying of the forests might be a short-term, not a long-term trend.

But even if the past ten years were peculiar and temperatures drop in the future, the trees may not return to their previous levels of production, warns Juday. “There are all sorts of thresholds built into these systems that are responding, and when you cross those thresholds, you can not confidently expect it [the forest] will come back. It may not. In fact, it very often does not.”

In Siberia, for example, as the Ice Age ended about 20,000 years ago, forests expanded up to the Arctic shore as ice sheets retreated. About 8,000 years ago, tree rings, ice and sediment cores, and other paleoclimate records show that temperatures began to drop. (The cooling may have been triggered by a massive flood of glacial meltwater into the northern Atlantic Ocean.) Despite cooler temperatures, the forest continued to grow until the eruption of the Mediterranean volcano Santorini around 1628 BC. The sulfur aerosols that the eruption pumped into the atmosphere further cooled temperatures for a few years, and during those years, the northernmost forests died. Though warmer temperatures returned, the trees never came back, says Juday. Tundra replaced the forest.

Researchers studying tree growth rings confirm that trees in the boreal forest don’t necessarily grow more vigorously in warm temperatures. In fact, the growth of 40% of the trees sampled slowed when July temperatures rose above 16.5°C (Photography copyright Glenn Patrick Juday, University of Alaska Fairbanks.)

The current boreal forest would likely be replaced with other types of forest as climate warms, but as the forest changes, its role in the carbon cycle could be reversed: it could turn from being a carbon sink to being a carbon source. Currently, the trees tuck away carbon and eventually transfer it to the frozen soil, where carbon accumulated over the past 10,000 years is now stored. “But if you turn your refrigerator off, things decay,” Houghton explains by analogy, and the decaying soil would release carbon dioxide. The trees themselves could also become a source of carbon dioxide if fires or insects kill large swaths of forest. Thawing, burning, and insect outbreaks are increasingly probable as temperatures rise.

Only time will tell if the drop in productivity measured by satellites and confirmed by ground-based carbon dioxide observations is a symptom that the boreal forests have reached a critical threshold. Angert observes, “I guess most people believe that there will be some turning point, but they believe it will be a few decades ahead when it gets so hot that the biosphere starts emitting carbon dioxide. These studies hint that we’re already close on this turning point.”

References:
• Angert, A., Biraud, S., Bonfils, C., Henning, C.C., Buermann, W., Pinzon, J., Tucker, C.J., and Fung, I. (2005) Drier summers cancel out the CO² uptake enhancement induced by warmer springs. Proceedings of the National Academy of Sciences, 102(31), 10823-10827.
• Arctic Climate Impact Assessment. (2004). Cambridge University Press. http://www.acia.uaf.edu/
• Goetz, S.J., Bunn, A.G., Fiske, G.J., and Houghton, R.A. (2005) Satellite-observed photosynthetic trends across boreal North America associated with climate and fire disturbance. Proceedings of the National Academy of Sciences, 102(38), 13521-13525.

Related Links:
• Introduction to BOREAS, the Boreal Ecosystem-Atmosphere Study from NASA’s Earth Observatory
• The Migrating Boreal Forest from NASA’s Earth Observatory
• Global Garden Gets Greener from NASA’s Earth Observatory
• Global Inventory Modeling and Mapping Studies (GIMMS)
• Boston University’s “The Greening Earth” Page (Includes bibliography of related research.)

Ancient tree stumps and logs dot the tundra north of the treeline in Siberia. Temperature records derived from the fossil wood suggest that a cold period associated with volcanic aerosols destroyed the forest in 1628 BC. Even after warmer temperatures returned, the forest never recovered. (Photograph copyright Jan Esper Swiss Federal Institute for Forest, Snow and Landscape Research.)