Lorey´s height regression for ICESAT-GLAS waveforms in hyrcanian deciduous forests of Iran

Since Lidar technology provides the most direct measurements of 3D of phenomena, it plays a critical role in a variety of applications. Forest canopy height as a main factor in forest biomass estimation is costly and time consuming to be measured on the ground. This study aims to estimate Lorey´s height “H_lorey” using GLAS data based on regression models. Different metrics like waveform extent “W_ext”, trail-edge extent “Htrail” and lead-edge extent “H_lead” were extracted from waveforms and a terrain index “TI” was also calculated using a digital elevation model. H_lorey estimated using multiple regression models were compared to field measurements data. A 5-fold cross validation method was used to validate the results. Best model with lowest AIC (297.440) was resulted using combination of W_ext and TI (R_a²=0.72; RMSE= 5.04m). The results show capability of ICESat-GLAS to estimate Lorey´s height in sloped area with a simple regression model. It is prospected to reach better result using other statistical methods and also improvement of processing techniques for LiDAR waveforms in the case of sloped terrain.