You are here: Home / Research Topics / Research Highlights / Individual Highlight

Research Highlights

Individual Highlight

A Tree Level Model of Forests in the Western United States

Photo of A subset of the landscape in Montana’s Swan Valley (top panel). The lower panel shows the plot IDs for the best-matching plot for each pixel of the same landscape, with each color representing a unique plot. In the left half of the imagery, the landscape is dominated by a checkerboard pattern, the legacy of extensive timber harvest on private lands, and less extensive harvest on public lands. On the right side of the imagery, vegetation is dominated by topographic gradients in a mountainous landscape. The model was able to pick up these patterns, with the outline of the checkerboard visible in the left half of the lower panel, and the topographic gradients visible in the clustering of the plots on the right half of the panel. U.S. Department of Agriculture Forest Service.A subset of the landscape in Montana’s Swan Valley (top panel). The lower panel shows the plot IDs for the best-matching plot for each pixel of the same landscape, with each color representing a unique plot. In the left half of the imagery, the landscape is dominated by a checkerboard pattern, the legacy of extensive timber harvest on private lands, and less extensive harvest on public lands. On the right side of the imagery, vegetation is dominated by topographic gradients in a mountainous landscape. The model was able to pick up these patterns, with the outline of the checkerboard visible in the left half of the lower panel, and the topographic gradients visible in the clustering of the plots on the right half of the panel. U.S. Department of Agriculture Forest Service.Snapshot : Maps of the number, size, and species of trees in forests across the western U.S. are desirable for a number of applications including estimating terrestrial carbon resources, tree mortality following wildfires, and for forest inventory; however, detailed mapping of trees for large areas is not feasible with current technologies. Forest Service scientists used a statistical method called random forests for matching forest plot data with biophysical characteristics of the landscape to populate entire landscapes with a limited set of forest plot inventory data.

Principal Investigators(s) :
Riley, Karin 
Research Location : Western United States, Arizona, California, Colorado, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington, Wyoming
Research Station : Rocky Mountain Research Station (RMRS)
Year : 2016
Highlight ID : 1150

Summary

Forest Service scientists sought to match plot data collected by the agency’s Forest Inventory Analysis (FIA) with characteristics of the landscape, as mapped on a 30 meter by 30 meter (98 feet by 98 feet) grid by the LANDFIRE project. The result is a map of plot identifiers, with the identifier of the best-matching plot assigned to each grid cell. The scientists used a modified random forests approach, which utilizes a “forest” of decision trees in order to identify the best-matching plot for each grid cell. One of the strengths of the random forests method is that it can model complex nonlinear relationships. map of plot identifiers can be linked to FIA’s databases to produce tree-level maps (aka “tree list”) or to map a number of other plot attributes. For example, they used the map of plot IDs to generate maps of forest cover, forest height, and existing vegetation group at 30 meter x 30 meter resolution for all forested pixels in the western United States.

They found high levels of agreement between our dataset and LANDFIRE data indicate our modified random forests model was able to identify forest plots that closely matched the landscape characteristics of 30 meter x 30 meter grid cells.

As the dataset in essence provides a tree-level model of forests in the western U.S., it greatly augments the information available to researchers and managers who previously relied on data from sparse forest plots or stand inventories.