Record Details
Spatially-explicit prediction of recoverable harvesting residues
ScholarsArchive at Oregon State University
| Field | Value |
|---|---|
| Title | Spatially-explicit prediction of recoverable harvesting residues |
| Names |
Schmidt, Christian F. W.
(creator) Sessions, John (advisor) |
| Date Issued | 2011-07-01 (iso8601) |
| Note | Graduation date: 2012 |
| Abstract | In order to decide whether or not to develop biomass energy facilities and where to best locate them, developers and investors need accurate assessments of fuel supply. This includes information about the distribution and concentration of fuel throughout the assessment area, the quality of fuel (form, moisture content, contaminant content, energy value, etc.), accessibility, and transportation distances to the facility. Including spatial detail about recoverable fuel densities and distribution allows planners to determine optimal choices for facility siting, fuel sourcing and transportation. Improving assessment accuracy increases the likelihood of project success and profitability. In this thesis we demonstrate methods for predicting quantity, density and distribution of recoverable forest harvesting residues applied to a biomass assessment project in northwest Oregon and southwest Washington. Predicted vegetation maps, allometric equations, and information about landowner intentions were used to predict forest biomass during the inventory period. Riparian and wetland management zones were modeled in a Geographic Information System (GIS) and excluded from potential biomass yielding lands. Private industrial forests harvestable during the inventory and forecast periods were classified into groups by species composition and basal area. To represent growth between inventory and forecast periods average biomass densities for inventory period harvestable forest groups were projected onto the groups in forecast period harvestable forests. On Oregon Department of Forestry (ODF) land, harvest plan data was used to determine harvest areas in the forecast period and how much biomass they would yield. This information was combined to make a GIS raster of recoverable harvest residue densities for the forecast period. The estimate of residues recoverable over the forecast period was 7,054,526 bone-dry tons (BDT), using the strictest residue criteria. A simulation was designed to quantify uncertainty caused by use of allometric equations in the inventory period private industrial forest biomass estimate. This uncertainty simulation generated an average recoverable biomass total of 6,412,049 BDT, only 0.4% less than the point estimate of 6,437,632 BDT. The maximum and minimum of 1170 simulation runs differed by only 392 BDT. This suggests that the allometric equations have a negligible contribution to the uncertainty of the total biomass estimate. Comparison of the point estimate, simulation outputs and rough estimates of biomass from harvest records and estimated BDT/MBF ratios led us to conclude the point estimate for the inventory period was reasonably accurate. This lent support to the forecast period estimate which relied on many of the same methods. Calculated average BDT/MBF ratios for the project area were comparable to other published figures. Promising subjects for future research include improving the gradient nearest neighbor (GNN) imputation method to predict recoverable biomass, developing methods for quantifying uncertainty of biomass assessments, and determining residue recovery rates for different forest types and harvest methods. |
| Genre | Thesis/Dissertation |
| Topic | logging residues |
| Identifier | http://hdl.handle.net/1957/21928 |
