Simulating the effects of lidar scanning angle for estimation of mean tree height and canopy closure

Abstract

A three-dimensional simulation model was used for modeling the scanning angle effect when measuring tree height and canopy closure in boreal forest with a laser scanner. The height distribution of the laser returns and the proportion of laser returns from the canopy were simulated using ray-tracing applied to a computer modeled forest. The proportion of canopy returns is commonly used as a measure of canopy closure, and height percentiles are commonly used to estimate mean tree height. Laser scanner data and field measurements of tree position, tree height, crown diameter, and crown base height were used for validating the simulation model. The correlation coefficient between simulated and real laser height percentiles was 0.96 and the simulation model systematically overestimated the laser height percentiles by 2.25 m. Simulations show that laser height percentiles and proportion of canopy returns changed more with an increased scanning angle for long crown species like spruce, compared with short crown species like pine. The change of height percentiles due to scanning angle was greater in forests with low stem numbers than with high stem numbers. The proportion of canopy returns was more affected by scanning angle than were the laser height percentiles.