Using LiDAR and GIS to extrapolate data from a small watershed to watershed-scale to provide insight into patterns and relationships in an Oregon central-western Cascade forest
Abstract
The combination of LiDAR (Light Detection and Ranging), data from field plot measurements, and GIS technology can provide new insights into spatial variation in ecological patterns and processes. We took advantage of long-term field data in a small (96 ha) watershed (“Watershed 1”) in the H.J. Andrews Experimental Forest, an LTER (Long Term Ecological Research) site situated in Oregon’s central-western Cascade Range. Originally an old-growth Douglas-fir forest, the entire basin was clearcut in the late 1960s and then planted with Douglas-fir seedlings. However, three other tree species have found their way into the watershed and with highly variable terrain and nutrient availability, vegetation growth is also highly variable. A network of 133 vegetation plots was installed to track subsequent growth and development of vegetation and has been re-measured approximately every 5-8 years ever since. A wealth of other information is available for the site thanks to decades of intensive research. This information has been incorporated into GIS layers along with high-resolution LiDAR measurements acquired in 2008. Our goal was to use the LiDAR data to extrapolate data from plot measurements to the entire basin, and then to use the derived images to evaluate landscape patterns and relationships. The slopes of the watershed are very steep in many places, and the field plots were originally laid out as fixed-radius circles with no slope correction. We obtained significantly better correlations between field plot data and LiDAR data when the dimensions of each plot was corrected according to the local slope and aspect, resulting in variable-sized ellipses. We explored various combinations of LiDAR metrics related to canopy cover and height, and explored various linear and non-linear fittings for combinations of LIDAR metrics. We found that we could explain approximately 50% of the observed variation in vegetation properties such as biomass and productivity with LiDAR. Relationships between vegetation productivity and other site variables are currently being explored.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2010
- Bibcode:
- 2010AGUFM.B41A0296Q
- Keywords:
-
- 0428 BIOGEOSCIENCES / Carbon cycling;
- 0439 BIOGEOSCIENCES / Ecosystems;
- structure and dynamics;
- 0466 BIOGEOSCIENCES / Modeling