LiDAR data are increasingly available and have a wide range of applications, from generating digital elevation models and assessing archeological resources to estimating forest biomass. LiDAR specifications vary widely depending on the purpose of the acquisition; terrain applications are generally flown at 1 pulse per square meter, while 7-8 pulses per square meter are preferred for forestry work. We investigated the suitability of leaf-off moderate density LiDAR data (mean density 3.8 pulses per square meter, acquired for an ecological land typing project) for assessing basic forest structural classes. Our primary objective was to determine if these data could be used to identify areas likely to contain early successional habitat, which is generally lacking across the landscape of northern New England. Our study area is a small (6880 ha) watershed on the White Mountain National Forest in New Hampshire, USA, characterized by northern hardwood forests at lower elevations transitioning to spruce-fir at higher elevations. LiDAR returns were processed using FUSION software to generate a variety of canopy metrics related to height, cover, and density; these were converted to raster format for analysis in ArcMAP.
Results/Conclusions
One product that is showing considerable utility in early testing is a canopy height model with 1 meter resolution. The map can be used to classify areas by canopy height, but also provides spatial information on past forest treatments and disturbance. Canopy openings, even of small sizes, are clearly visible years after they occur, even though the data were acquired during leaf-off conditions. We validated the canopy height model with aerial NAIP imagery and found good agreement between the two in regard to canopy openings, with the canopy height model providing greater spatial resolution. Another early product derived from the LiDAR metrics is a canopy cover map (20m pixel) classified into three categories: >70% canopy, 30-70%, and <30%. The 30% canopy cover value is a cutoff that often characterizes early successional habitat in these forests. Early validation shows good agreement between bird species observed at study plots and the cover classification of those plots as modeled with the LiDAR data. Ongoing work includes continued validation, development and validation of additional map products, such as understory density, and application and testing of the models on additional areas across the White Mountain National Forest for which LiDAR data are available.