Evaluation and testing of standardized forest vegetation metrics derived from lidar data

Science Center Objects

The USGS 3D Elevation Program (3DEP) is managing the acquisition of lidar data across the Nation for high resolution mapping of the land surface, useful for multiple applications. Lidar data is initially collected as 3-dimensional “point clouds” that map the interaction of the airborne laser with earth surface features, including vegetation, buildings, and ground features. Generally the product...

The USGS 3D Elevation Program (3DEP) is managing the acquisition of lidar data across the Nation for high resolution mapping of the land surface, useful for multiple applications. Lidar data is initially collected as 3-dimensional “point clouds” that map the interaction of the airborne laser with earth surface features, including vegetation, buildings, and ground features. Generally the product of interest has been high resolution digital elevation models generated by filtering the point cloud for laser returns that come from the ground surface and removing returns from vegetation, buildings, powerlines, and other above ground features. However, there is a wealth of information in the full point cloud on vegetation structure that is currently being underutilized in USGS 3DEP product delivery. Normalizing the elevations of above ground objects to height above ground allows for characterizing the 3D nature of vegetation with lidar data and enables mapping vegetation height variability, structure, and volume over large areas. These mapped attributes have proven to be extremely useful for habitat studies, vegetation biomass and biomass change studies, and wildfire behavior models. In this project we developed methods to standardize vegetation metric geospatial product generation from USGS 3DEP point clouds using scripting in Python and R, we generated standardized metric products for two National Park pilot areas from five recent lidar acquisitions, and we developed a web front end to demonstrate product search, selection, and download of merged processed tiles at two spatial resolutions. Possible next steps include development of near real-time, on-the-fly processing of data stored in the cloud for individualized product delivery.