Event Website
http://www.nafew2009.org/
Start Date
6-23-2009 11:50 AM
End Date
6-23-2009 12:10 PM
Description
The goals for a wide range of forest management objectives are often stated in terms of the amount and layering of canopy cover. However, measuring canopy cover is labor intensive and different techniques provide widely different estimates. Several approaches have been developed to predict cover from common tree or stand-level density attributes, with varying results. This study used line-intercept measured tree cover from 1,424 Forest Inventory and Analysis (FIA) plots across Oregon to build predictive models from estimates of tree stocking, crown width, and other stand attributes (mean diameter, stand height, SDI, etc.). A variety of adjustments were applied to adjust for tree social status and account for tree crown overlap. Stocking was a better predictor of cover than crown width, although much of the error in the latter was due to estimates of crown overlap. The random crown overlap function that is standard in the Forest Vegetation Simulator (FVS) resulted in biased predictions in mesic forest types, but not in dry forest types. New model predictions based on stocking for mesic forest types were within 15 percent of measured cover for > 82% of the observations. Although there are some additional options for improving estimated cover from tree and stand attributes, ground-based measurements will probably be required for precise estimates.
Included in
Canopy Cover Prediction From Stand Density Attributes: Stocking, Crown Width, and Overlap Functions
The goals for a wide range of forest management objectives are often stated in terms of the amount and layering of canopy cover. However, measuring canopy cover is labor intensive and different techniques provide widely different estimates. Several approaches have been developed to predict cover from common tree or stand-level density attributes, with varying results. This study used line-intercept measured tree cover from 1,424 Forest Inventory and Analysis (FIA) plots across Oregon to build predictive models from estimates of tree stocking, crown width, and other stand attributes (mean diameter, stand height, SDI, etc.). A variety of adjustments were applied to adjust for tree social status and account for tree crown overlap. Stocking was a better predictor of cover than crown width, although much of the error in the latter was due to estimates of crown overlap. The random crown overlap function that is standard in the Forest Vegetation Simulator (FVS) resulted in biased predictions in mesic forest types, but not in dry forest types. New model predictions based on stocking for mesic forest types were within 15 percent of measured cover for > 82% of the observations. Although there are some additional options for improving estimated cover from tree and stand attributes, ground-based measurements will probably be required for precise estimates.
https://digitalcommons.usu.edu/nafecology/sessions/sdi/5