Estimating woody biomass feedstock across a landscape consists of three basic steps: (1) quantifying estimates of forest characteristics, such as basal area, trees, and woody biomass tons per acre across a landscape; (2) using those estimates to help determine where to apply actual or hypothetical silvicultural prescriptions; and (3) combining estimates of woody biomass with prescriptions to calculate potential treatment residues that can be utilized for fuel or raw material. Quantifying existing forest characteristics can be a substantial endeavor. Generally, this process consists of sampling areas on the ground and recording tree measurements, such as species counts, diameter at breast height (1.37 m), total height, live crown ratio, age, and percentage cull and breakage [15]. From these tree measurements, estimates of standing volume and weight are calculated using allometric equations. These measurements and calculations are then summarized based on sampling design to describe multiple aspects of a forest on a per acre basis. A common classical approach to quantifying existing forest characteristics uses stratified random sampling to relate summarized values to polygons within groups (strata) of similar forest types, stockings, and canopy cover [16]. With this approach, polygons and strata are generally created and labeled through manually defining boundaries of similar forest cover types, percentage canopy cover, and topographic position derived from aerial and satellite imagery. For larger landscapes where manual interpretation is impractical, image classification techniques are used to develop appropriate strata. Once strata have been defined, a random sample of polygons within each stratum is selected, visited, and sampled to derive mean estimates of forest charac­teristics for that stratum. Mean strata estimates are then attributed to each polygon within each stratum.

While this basic approach is still used in many analyses, mean estimates relate to the stratum as a whole and do not account for spatial variations within a given stratum. Fur­thermore, the coarse grain nature of this type of estimate may not be suitable for fine scale projects that utilize only small portions of strata. To address this issue, recent anal­yses have developed spectral and textural relationships between remotely sensed data and field measurements [17-19]. Using these relationships, estimates of biomass can vary as spectral and textural values change, thereby maintaining the spatial heterogeneity of forest characteristics at fine spatial resolution across the landscape.

After forest characteristics have been quantified for polygons or cells, they can be used to help determine where silvicultural prescriptions are applied across a landscape. The process of allocating these prescriptions to forested areas can be done in a similar man­ner as allocating logistical cost. Specifically, rules can be developed and applied using the attributes of spatial objects to identify polygons, portions of polygons, or cells that meet defined thresholds. Once allocated, these prescriptions can be combined with quan­tified forest characteristics to provide spatially explicit estimates of potential total woody biomass that can be removed from a given location. Finally, depending on the efficacy of the harvesting system and the merchandizing of the trees, treatment residues can be calculated for a given location. These residues represent the amount of potentially avail­able woody biomass that can be utilized for energy and incorporated into potential woody biomass flows.