Remote Sensing Applications of a Stem-Map Model
for Predicting Canopy Cover of FIA Plots

Tree canopy cover is an important stand characteristic affecting understory light, fuel moisture, decomposition rates, wind speed, and wildlife habitat. Spatial products depicting per-pixel estimates of canopy cover can be developed from photo-interpreted reference data, but could benefit from the availability of georeferenced field measurements of tree canopy. A model is described for predicting canopy cover of USDA Forest Service Forest Inventory and Analysis (FIA) plots by mapping the locations of trees = 5.0 in. diameter within the plot, and statistical modeling of the sapling contribution to total cover. The model was designed with an operational focus, including the requirement that it scale efficiently to national applications. Model predictions were evaluated for use in remote sensing by generating LANDSAT-based prototype canopy cover maps for two study areas, one in the southwestern U.S. and one in the Northern Rockies. The prototype maps validate comparatively well against direct field measurements of tree canopy, and exhibit characteristics desirable in fire behavior modeling.