Remote Sensing Protocols for Parameterizing an Individual, Tree-Based, Forest Growth and Yield Model
Abstract
Potential impacts of climate change to southeastern U.S. pine ecosystems are of particular concern to the Department of Defense. The U.S. Forest Service-developed Forest Vegetation Simulator-Southern Variant (FVS-sn) forest growth model can project growth in southeastern U.S. pine ecosystems, and it has been modified to incorporate the effects of climate change. Stand inventories are typically utilized to parameterize FVS sn growth models, but field-based inventories are cost-prohibitive to collect at landscape scales. Therefore, remote sensing protocols were developed to parameterize the FVS-sn model. More specifically, a tree-finding model was developed to estimate the location and height of individual stems using LIDAR data. Estimated stem locations from the tree-finding model matched 74% and 98% of field-mapped longleaf and loblolly stems, respectively. Using estimates of stem height, height to live crown, localized stem density, and crown area for a total of 160 matched stems as predictor variables in regression analysis explained 68% and 71% of the variation in field-measured diameter at breast height (dbh) for longleaf and loblolly stems, respectively. Using this protocol, a landscape-wide map of stem locations attributed with species, height, dbh, and crown length could then be used to parameterize the FVS-sn model.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 2014
- Accession Number
- ADA609386
Entities
People
- George Z. Gertner
- Patrick J. Guertin
- Scott A. Tweddale
Organizations
- Engineer Research and Development Center