BLM REA YKL 2011 YKL DV C Alaska ForestBiomass.img

Some of the YKL rasters intentionally do not align or have the same extent. These rasters were not snapped to a common raster per the authors' discretion. Please review selected rasters prior to use. These varying alignments are a result of the use of differing source data sets and all products derived from them. We recommend that users snap or align rasters as best suits their own projects. - An aboveground live forest biomass map for the conterminous U.S., Alaska and Puerto Rico is derived from modeling field biomass estimates, collected nationwide by the USDA Forest Service Forest Inventory and Analysis (FIA) program, as functions of 250-m resolution satellite image products and other digital geographic layers. These predictor layers included the following: 16-day Moderate Resolution Imaging Spectrometer (MODIS) composites, associated vegetation indices, and percent tree cover; vegetative diversity and type synthesized from the National Land Cover Dataset (NLCD); topographic variables; monthly and annual climate parameters; and other ancillary variables. We segmented the U.S. into 65 ecologically similar mapping zones, plus Alaska and Puerto Rico. Before modeling biomass, inventory data served as the basis for classifying the predictor layers into a forest mask with the nonparametric classifier, See5©. Forest biomass models within the predicted forest areas used tree-based algorithms in Cubist©. Using independent test data, the estimated proportion of correctly classified pixels for the forest mask ranged from 0.85 in the Pacific Northwest to 0.94 in Alaska, while estimates of Kappa ranged from 0.57 in Puerto Rico to 0.88 in Alaska. For biomass, the largest model correlation coefficients between observed and predicted values, of 0.66 to 0.78, occurred in the Pacific Northwest and Interior West, while model correlation coefficients were smaller, with most below 0.40 in the eastern mapping zones. Design- and map-based estimates of total forest area and total aboveground live forest biomass are compared for individual states as well as four scales of spatial aggregation. An estimate of C pools in live forest biomass of U.S. forests, derived from the nationwide biomass map, is also compared to previously published estimates. This article documents the national geospatial predictor layer database, standardizing the national FIA data, developing predictive models, producing the maps with accompanying map uncertainty, and assessing model errors.