Estimating forest structural characteristics with airborne lidar scanning and a near-real time profiling laser systems

LiDAR (Light Detection and Ranging) directly measures canopy vertical structures, and provides an effective remote sensing solution to accurate and spatiallyexplicit mapping of forest characteristics, such as canopy height and Leaf Area Index. However, many factors, such as large data volume and high costs for data acquisition, precludes the operational and practical use of most currently available LiDARs for frequent and large-scale mapping. At the same time, a growing need is arising for realtime remote sensing platforms, e.g., to provide timely information for urgent applications. This study aims to develop an airborne profiling LiDAR system, featured with on-the-fly data processing, for near real- or real- time forest inventory. The development of such a system involves implementing the on-board data processing and analysis as well as building useful regression-based models to relate LiDAR measurements with forest biophysical parameters. This work established a paradigm for an on-the-fly airborne profiling LiDAR system to inventory regional forest resources in real- or near real- time. The system was developed based on an existing portable airborne laser system (PALS) that has been previously assembled at NASA by Dr. Ross Nelson. Key issues in automating PALS as an on-the-fly system were addressed, including the design of an archetype for the system workflow, the development of efficient and robust algorithms for automatic data processing and analysis, the development of effective regression models to predict forest biophysical parameters from LiDAR measurements, and the implementation of an integrated software package to incorporate all the above development. This work exploited the untouched potential of airborne laser profilers for realtime forest inventory, and therefore, documented an initial step toward developing airborne-laser-based, on-the-fly, real-time, forest inventory systems. Results from this work demonstrated the utility and effectiveness of airborne scanning or profiling laser systems for remotely measuring various forest structural attributes at a range of scales, i.e., from individual tree, plot, stand and up to regional levels. The system not only provides a regional assessment tool, one that can be used to repeatedly, remotely measure hundreds or thousands of square kilometers with little/no analyst interaction or interpretation, but also serves as a paradigm for future efforts in building more advanced airborne laser systems such as real-time laser scanners