Fire, logging, insects and extreme weather can wreak havoc on forests. With support from the Department of Energy and the National Science Foundation, Oak Ridge National Laboratory climate modeler Daniel Hayes and University of Tennessee–Knoxville ecologist Joseph Hughes map eastern U.S. forests to characterize changes. They use satellite imagery collected by the National Aeronautics and Space Administration as well as ground data from the U.S. Department of Agriculture Forest Service and the U.S. Geological Survey to calibrate and validate the satellite data. The maps help improve the models that are set in motion in simulations to explore the effect of landscape alterations on the carbon cycle. Hughes will present the research Aug. 7 in Minnesota during a talk at the annual meeting of the Ecological Society of America.
Left: Daniel Hayes, shown here outside of Nome, traveled to the Arctic in June to study climate change. Image credit: Santonu Goswami
“We want the models to capture all the important processes that impact the carbon cycle,” said Hayes, a research scientist in ORNL’s Climate Change Science Institute and Energy and Environmental Sciences Directorate. He holds a joint appointment as an assistant professor in UTK’s Department of Ecology and Evolutionary Biology, where Hughes is his graduate student. “Landscape changes—from turning a forest into a Walmart parking lot to insect outbreaks and fires—have implications for climate.”
Current Earth system models do not represent land disturbances in much detail, if they represent them at all, according to Hayes. He is interested in exploring feedbacks from such disturbances as forest fires to the carbon cycle, energy balance and hydrology.
The researchers rely on remote sensing data from Landsat, a program jointly managed by NASA and USGS. Since 1972, the program’s satellites have collected high-resolution imagery of Earth. “Every 16 days we get a new picture of the same spot,” said Hughes, who maps data collected between 1984 and 2011 at a typical resolution of 30-meter-wide pixels. “That’s an amazing time series and really great data density. You can’t see a car, but you can see roads and streams and forests.”
The researchers use Thematic Mapper sensors of Landsats 4, 5 and 7 to measure signatures of land, ice, clouds and more. Plants reflect mostly in the green bandwidth of visible light and the near-infrared bandwidth. Two mid-infrared bandwidths, which transmit through the sky with less scattering than do longer wavelengths, help the researchers screen out atmospheric noise. A long-wave infrared bandwidth measures temperature.
The images capture change that is often difficult to witness, such as insect infestations that damage forests over several years. “We need lots of data and pretty sensitive methods to find these slow, creeping changes,” said Hughes, whose project, sponsored by NSF’s Coupled Human/Natural Systems program, generates disturbance maps that have helped scientists evaluate the effectiveness of protecting areas such as those managed by the Nature Conservancy.
Left: A map of total forest changes near Newport, Tennessee, from 2000 to 2011 shows red regions that experienced vegetation removal (due to fire or logging), blue areas of regrowth, and green zones that are a mixture of both. Darker areas experienced more aggregate change than lighter areas, such as the Cherokee National Forest to the southeast and Douglas Lake to the west. Image credit: Joe Hughes
Hayes uses the temporally and spatially explicit dataset captured in Hughes’s forest maps to quantify disturbances in models and discern their dynamics with the climate. His models track carbon as it cycles through land, oceans, ice and atmosphere. Among land disturbances, fire has the biggest effect on the global climate budget. Combustion quickly sends carbon to the atmosphere as carbon dioxide or methane. Its lasting effects reset forest patterns for regrowth, carbon uptake, energy balance and water movement.
“We know fire is extremely important because models that don’t incorporate fire don’t match the reality of the global carbon cycle very well,” Hayes said. “You not only need to be able to have the fire in the model when it happens, you need to have fires included in your model somehow in the past because everything that happens in the past has a legacy.”
Hayes is figuring out how to best represent past events to capture legacy effects. As global warming has made some areas warmer and drier, fires in those regions have become more frequent and severe. “It’s certainly happening in the boreal regions,” Hayes said. “A huge area of unmanaged forests has big implications for climate.”
Hayes will next expand the geographic scope of techniques he and Hughes developed for eastern forests. For DOE’s Next Generation Ecosystem Experiments project in the Arctic, on which Hayes collaborates, his models may shed light on major climate feedback processes and reduce uncertainty in climate prediction. Such efforts are urgently needed because, as a result of amplification of natural responses to human-caused global warming at high latitudes, the Arctic is warming twice as fast as the rest of the planet.–by Dawn Levy Posted August 1, 2013 12:40 p.m.
Hughes: Wednesday, August 7, 2013: 8:40 AM
L100B, Minneapolis Convention Center