Rapid Research on a Record-Setting Forest Fire

For his PhD fieldwork, Peter Wolter spent the summers of 2006 and 2007 living in his Subaru Outback in and around the Boundary Waters Canoe Area Wilderness of northeastern Minnesota. With a background in remote sensing, his goal was to “ground truth” satellite images of the area as part of a project on the dynamics of the spruce budworm. He hiked a couple hundred miles of the sub boreal forest, noting the species, height, canopy diameter, basal area, and other characteristics of more than 10,000 trees on 120 plots. He used the field data and partial least-squares regression to identify and integrate the most meaningful variables from Earth-monitoring satellites such as Landsat and other sensors. The result was an extremely robust model of the forest structure and tree species in 3600 square kilometers. Published in Remote Sensing of Environment, the work pushed the state of the art in validating satellite imagery, says Phil Townsend, Wolter’s advisor and a forest ecologist at the University of Wisconsin. In fact, the resulting fine-tooth-comb analysis of the forest went well beyond what Townsend needed for analyzing the budworm’s host species. “The joke was,” Wolter recalls, “I have all this great data; now what are we going to do with it?” The answer came unbidden last year. On 18 August, lightning struck in the Boundary Waters near Pagami Creek. It started a bog fire that smoldered for about a week. Then humidity dropped in already-dry conditions as winds picked up, carrying the fire into treetops and creating an inferno. On 12 September, pushed by gusts reaching 35 miles per hour, the flames consumed roughly 60,000 acres of birch and jack pine in some five hours. Smoke from the firestorm drifted through Chicago and eventually to Europe. By the time it was contained in late September, the fire had torched 93,000 acres, setting a record as the country’s largest crown fire east of the Mississippi in more than 100 years. Wolter, who at the time was finishing a postdoctoral position with Townsend and on his way to a new job at Iowa State University, could not help watching the daily reports of the growing blaze. He pulled up maps and Landsat images of the fire and noticed something compelling: The burning Boundary Waters forest sat entirely within a corner of the square he had carefully modeled. Six of the 120 plots he had surveyed were going up in flames. Wolter’s dissertation data suddenly had a new purpose: as a prefire, finescale picture of the forest. Could a similar analysis after the fire explain which factors played into its severity? He took the idea to Townsend, whose initial reaction was “no way,” given the demands of the budworm research and other projects. But when he looked at the maps, his reaction changed to “no way we can’t do this.” They had data on the species and structure of the burned area with “100 percent coverage,” says Wolter. Adds Townsend, “we just put too much into that study area to let it go.” The task then became finding a way to study the dramatic ecosystem change in progress. Thanks to a Rapid Response Research award from the National Science Foundation, by 13 October, the scientists had secured funding, and with the cooperation of the US Department of Agriculture Forest Service, they assembled a crew and a field plan. They also arranged for spectroscopic imaging of the burnscape to assess its chemical composition: The National Aeronautics and Space Administration sent one of its planes with an airborne visible and infrared imaging spectrometer north from the oil-contaminated Gulf of Mexico to fly over the Boundary Waters burn. On 24 October, the field crew paddled into the zone of the contained but still active fire. They found a blackened land with fuel from a century of fire suppression completely consumed. The fire, says Wolter, “burned all the organic matter down to mineral soil.” But the fire-adapted forest also showed signs of life: Yellow fungi sprouting on black char, seeds dispersed from jack pine cones unsealed by the fire, wolf prints everywhere. The crew carried out hundreds of pounds of soil that is being analyzed for carbon and nitrogen, as well as mercury, to better understand how fire intensity and forest type interact to affect movement of the pollutant locally and globally. All that is a baseline for the fieldwork that is happening this summer, to capture the forest’s response to the fire during the first growing season and to see how the soil begins to regenerate. As large fires like the one at Pagami Creek return to the landscape, says Townsend, the episode serves as a natural experiment for understanding how forests respond to such dramatic change.