Arthur Temple College of Forestry and Agriculture

lidarlidarlidarlidarlidarlidar      Lidar derived digital terrain model with field
     sampling grid of the SFA Experimental ForestLiDAR (Light Detection and Ranging) technology has been applied in a wide variety of fields such as civil engineering, atmosphere monitoring, terrain mapping, and traffic control etc.  Based on the time a laser pulse travels between the sensor and the target, the position of an object can be precisely located in 3-dimentional (3D) space.  Due to the advancement of computer processing power and memory capacity, lidar applications have shown some promising features in forestry.  It is noted that European community has moved at a faster pace in using lidar on regular basis for forest inventory and management.  In the United States, forestry-lidar applications have just started to become commercialized.  Theoretically, forest biometrics at both individual tree level (total height and crown height, etc.) and stand level (volume, basal area and biomass, etc.) can be estimated from "lidar point cloud" through some kind of algorithm.  However, the accuracy remains unknown until ground measurement is conducted in a well planned manner.

The Stephen F. Austin Experimental Forest in east Texas is our research area for this project.  The first objective of our research is to conduct accuracy assessment by comparing ground measured data to lidar derived data.  Different combinations of data processing algorithms and data sources for lidar derivation will be used in order to see if one is significantly more accurate than another.  Eventually, a fine tuned model will be built as the best estimation of the forest through lidar remote sensing.

The second objective is to rastering lidar point cloud data for the fusion of multispectral imagery with the attempt to increasing cover type classification accuracy.  Large-area forest inventory relies on accurately delineating forest cover types through remote sensing.  Lidar derived raster data are expected to provide additional perspectives for distinguishing forest features due to its 3D structure.

Ultimately, we will integrate all of the data into a geodatabase where the position of each individual tree (x, y coordinate) and its associated attributes (species, DBH, height, basal area, crown diameter, etc.) are stored.  Through modeling, different forest practices such as thinning can be simulated before a decision is made.

Picture2Profile view of Lidar point cloud

pointCloudLidar point cloud of the SFA Experimental Forest

Picture1Reseaching at the SFA Experimental Forest