Derivation of forest inventory parameters for carbon estimation using terrestrial LiDAR

Abstract

This research was conducted to derive forest sample plot inventory parameters from terrestrial LiDAR (T-LiDAR) for estimating above ground biomass (AGB)/carbon stocks in primary tropical rain forest. Inventory parameters of all sampled trees within circular plots of 500 m2 were collected from field observations while T-LiDAR data were acquired through multiple scanning using Reigl VZ-400 scanner. Pre-processing and registration of multiple scans were done in RSCAN PRO software. Point cloud constructing individual sampled tree was extracted and tree inventory parameters (diameter at breast height-DBH and tree height) were measured manually. AGB/carbon stocks were estimated using Chave et al., (2005) allometric equation. An average 80% of sampled trees were detected from point cloud of the plots. The average of plots values of R2 and RMSE for manually measured DBHs were 0.95, 2.7 cm respectively. Similarly, the average of plots values of R2 and RMSE for manually measured trees heights were 0.77, 2.96 m respectively. The average value of AGB/carbon stocks estimated from field measurements and T-LiDAR manually derived DBHs and trees heights were 286 Mg ha-1 and 134 Mg ha-1 ; and 278 Mg ha-1 and 130 Mg ha-1 respectively. The R2 values for the estimated AGB and AGC were both 0.93 and corresponding RMSE values were 42.4 Mg ha-1 and 19.9 Mg ha-1 respectively. AGB and AGC were estimated with 14.8% accuracy.