|
Using Wave-specific Metrics to Estimate Forest Structure from Lidar Wave Data
Jordan
Muss, South Dakota State University, jdmuss@gmail.com
(Presenter)
Waveform lidar data have been used to estimate forest structure, including height, biomass, and basal area, but the models that have been created for these purposes have, overwhelmingly, used point data extracted from the waveforms. Two drawbacks to the decomposition of waveforms are: (1) a potential wealth of information that could be concealed within the waveform structure is disregarded; and (2) the quantile-based metrics used to summarize the point data are highly correlated. Recent studies have challenged these common approaches by introducing alternative methods to characterize waveform data, and have resulted in improved assessments of forest structure. We have adapted two distance measurements used in mechanical analyses to describe the manner in which the wave is distributed around a point. These metrics—the radius of gyration (RG) and moment distance index (MDI)—capture the shape of lidar waveforms around either the centroid of the wave (RG) or from user-defined points (MDI), and can be used with other wave-based metrics to improve estimates of forest structure. We have tested these metrics using pseudo-waves created from discrete return lidar data and forest measurements collected in northern Wisconsin. In many cases RG and MDI improved model fit, confirming that wave-based methods are robust and viable alternatives to frequency-based lidar analysis. Presentation Type: Poster Session: Science in Support of Decision Making (Wed 10:00 AM) Associated Project(s):
Poster Location ID: 199
|