Registration of multiple low resolution nasa airborne snow observatory (ASO) lidar data for forest vegetation structure caracterization

Airborne lidar is the tool best suited to provide timely updated maps for monitoring forest change in both horizontal and vertical dimensions. Still, it has been little used due to the scarcity of long-term time-series of lidar measurements. The NASA Jet Propulsion Laboratory Airborne Snow Observatory (ASO) is a landscape-level monitoring system that provides ongoing remote sensing measurements with high temporal resolution over large mountainous areas to quantify snow volume and dynamics. ASO collects multi-year low-resolution lidar data (∼1.5 pt/m²) with a nominal weekly frequency up to 12 times a year. In this work, we present a new method to automatically register ASO weekly low-resolution forest lidar point clouds in order to calculate spatially consistent datasets (∼18 pt/m²) adapted to fine scale forestry studies. Our method is tested using 12 lidar datasets acquired over the Tuolumne River Basin (California) in the spring and summer of 2014. On average, the ASO lidar system provides accurate measurements in terms of geolocation (0.38m and 0.12m for the horizontal and vertical dimension, respectively) but some datasets are biased up to 1.38m and 0.53m, respectively. Our registration method successfully corrected for systematic bias improving the 3D geometry of forest point clouds.