WWETAC ProjectsProject Title: Predicting forest inventory variables using lidar-derived metrics and FIA plot measurement data
Principal Investigator: Robert J. McGaughey, US Forest Service, PNW Research Station
Collaborators: Janet Ohmann, US Forest Service, PNW Research Station; Van R. Kane, University of Washington.
Key Issues/Problem Addressed:
Ideally, vegetation and fuels information to assist fire and fuel managers should: 1) integrate data on fuels, species composition, and forest structure in a common format, 2) be spatially explicit (i.e. mapped), 3) accurately portray fuels and vegetation with a small enough grain size to be useful for tactical (i.e. stand-level) decision making, 4) at the same time cover a large landscape or region to be useful for more strategic planning and decision making, and 5) provide information about the type, magnitude, frequency, and location of errors in each dataset. Unfortunately, information with all of these characteristics does not currently exist. The GNN vegetation maps that currently are available for most of the Pacific Northwest are based on Landsat satellite imagery and thus are appropriate for "mid-scale" applications in forest planning and policy, but are not considered sufficiently reliable for use in stand-level management. LIDAR-assisted forest inventory can provide detailed, spatially explicit information for overstory conditions but does not provide detail for lower canopy layers.
Setting and Approach:
In 2009 and 2010 high density LIDAR data were acquired for 1.9 million acres in central Oregon. The acquisition, known as the Deschutes Study Area was conducted as part of the ongoing campaign directed by the Oregon Department of Geology and Mineral Industries (DOGAMI). The acquisition covers the majority of the land managed by the Deschutes National Forest. Data were collected to produce high resolution, high accuracy bare earth models for use in hydrologic modeling and landslide mapping and risk assessment.
We propose to use data from recent FIA plot measurements and metrics computed from the Deschutes LIDAR data to model the relationships between the metrics and selected inventory variables (basal area, dominant stand height, stem volume, quadratic mean diameter, and stem density). Previous studies have shown that LIDAR can be used in conjunction with field plots to estimate similar inventory variables. In addition, we will process the entire LIDAR point cloud to produce a suite of descriptive metrics using cell sizes of 10, 30, and 90 m. These raster data layers, while useful by themselves, will be used with the predictive model to estimate the inventory variables across the entire acquisition area at a resolution of about ¼ acre (30 by 30 m cells).
Progress to Date:
LIDAR data for the Deschutes National Forest has been processed to produce descriptive metrics at resolutions of 10, 30, and 90m. These metrics, in the form of raster GIS data layers, have been delivered to cooperators at Oregon State University and PNWRS Corvallis (Dr. Janet Ohmann) for use in GNN imputation. FIA plot data (plots measured within 3-5 years of the LIDAR data acquisition) have been obtained, cleaned, and summarized. A portion of these were geo-referenced in 2010 and the remaining plots in 2011. Post-processing of the GPS data was completed by Deschutes NF staff early in 2012. Point cloud samples corresponding to the plots have been extracted and descriptive metrics for the samples computed for use in modeling inventory variables. The University of Washington cooperator, Dr. Van Kane, has begun the modeling work.
The GIS data layers derived from the LIDAR data and the predicted inventory information will provide information useful to a variety of resource specialists on the Deschutes National Forest. In addition, other researchers can use these layers to help impute additional variables describing forest structure and condition using a gradient nearest neighbor (GNN) approach. The resulting GNN maps will contain the standard suite of forest attributes as produced for the Interagency Mapping and Assessment Project (IMAP), as well as fuels-specific variables needed for fire modeling programs such as FARSITE and FlamMap. The addition of the LIDAR-derived information to the GNN approach should result in more spatially accurate GIS data layers which will, in turn, lead to better, more defensible management decisions.
WWETAC ID: FY11NG96