A GeOBIA-assisted framework for canopy height estimation using airborne lidar scanning: a city-scale approach

Airborne LiDAR scanning (ALS) provides a platform to remotely sense large areas of standing trees in a short period. Generally, scanning systems are mounted to fixed-wing aircraft with a nadir orientation that permits either single or multi-return data capture. Due to the discrete nature of a vector point cloud, spacing between individual points contains no data and has to either be ignored or interpolated based on the data points surrounding them. This could result in the underestimation of total stem height due to the apex of each stem falling within point spaces of zero data. In this study, a typical area of urban tree cover in Auburn, Alabama was examined to determine if underestimation is present and to what extent. Results showed an average total height underestimation of 6.6% across all size classes with higher rates of underestimation in shorter trees (<12 m) compared to the average underestimation of taller trees.