Estimating spatial forest stand variables such as mean height, mean crown diameter, mean diameter breast height DBH, tree density and aboveground biomass is important for sustainable forest management. This study aimed to estimate forest stand variables in coniferous tree species of Picea crassifolia stand in the Qilian Mountain, western China from single tree detection using small-footprint airborne LIDAR data. Based on the LIDAR data, a canopy height model (CHM) was firstly computed as the difference between tree canopy hits and the LIDAR terrain elevation values. In this study, a double-tangents crowns recognition algorithm was used to extract single tree location, height and crown polygon.
Stepwise multiple regression models were used to develop equations relating LIDAR-derived parameters, such as tree height, stand density and crown width, with observed forest parameters for each sample plot. The precision of equation for estimating mean stand height, tree density and aboveground biomass is high, with R2 bigger than 0.7. These results showed that the LIDAR data was useful for forest stand variables. Finally, the spatial forest stand variables maps were established using the stepwise multiple regression equations. The results showed that highdensity LIDAR data could be used to get forest variables distribution maps with relatively high precision, which was of important practical significance for sustainable forest management and update of forest form map, and for forest hydrological science research in small basin.