| dc.description.abstract |
Remote sensing tools that directly characterize canopy structure would be beneficial
for management activities and conservation planning. LiDAR (Light Detection And Ranging)
is such a tool, as an active remote sensing technology that provides fine-grained
information about the three-dimensional structure of ecosystems across a broad spatial
extent.
This project assesses the feasibility of using the state-wide North Carolina Floodplain
Mapping LiDAR dataset to differentiate between the structural components of evergreen
forest types in North Carolina’s coastal plain. Vertical structure and spatial patterns
of vertical structure were quantified using geospatial measures such as semivariogram/
correlograms, lacunarity analysis, and correlation length. LiDAR-derived metrics were
also created for comparison with standard field-based measurements of stand structure.
I found that LiDAR is capable of measuring canopy variation and can differentiate
between the structural characteristics of evergreen forest types. Also, the N.C. LiDAR
has potential for use as a surrogate for field measurements when collection is not
feasible due to time, labor, or financial constraints. In addition, the project examined
LiDAR’s use as a screening tool in the identification of suitable habitat for the
federally endangered red-cockaded woodpecker (Picoides borealis). I used Maximum Entropy
(Maxent), an inductive modeling algorithm for presence data only, to create a spatial
species distribution model using LiDAR-derived variables in addition to more typical
geospatial variables. The Area (AUC) under the Receiver Operating Characteristic (ROC)
curve was analyzed for increases in predictive power with additions of variables.
Results suggested that the addition of LiDAR-derived variables to habitat models improved
their predictive power, resulting in a test AUC increase from 0.923 with standard
spatial variables only, to a test AUC of 0.951 with LiDAR-derived variables added.
The success of this project has important implications for natural resource management
and conservation planning, especially given that the LiDAR dataset is publicly available
and covers the entire state of North Carolina.
|
|
