Linking Optical SPOT and Unmanned Aerial Vehicle data for a rapid biomass estimation in a Forest-savanna Transitional Zone of Ghana

Abstract

The direct estimation of biomass using remote sensing technologies, such as LiDAR, RaDAR and Stereo Data is limited in utility, since it does not allow for historical analysis of biomass dynamics far back in time due to their recency in development. This study links Unmanned Aerial Vehicle (UAV)-measured tree height and optical SPOT image reflectance in a mathematical model for a quick and less expensive indirect biomass estimation, and the possibility of historical analysis using the earliest captured optical data. SPOT 6/7 images were used to map land-use/cover patterns. A Phantom 4 drone images were used for height and crown width estimation. A stepwise regression analysis was conducted to establish a relationship between SPOT 6/7 channels and the UAV-generated tree heights. The linear model was used to convert the reflectance values of SPOT images into tree heights, and in turn used for crown width estimation. The estimated tree height and crown width images were used to estimate biomass using an allometric equation. There was no statistically significant difference between UAV and manual tree height measurements. UAV-estimated tree height predicted 88.0% of crown width. Regressing the tree height on the SPOT bands yielded an R² of 66.0%. It is recommended that further studies be conducted to improve on the accuracy of estimation. It is hoped this would facilitate a quick biomass estimation and long term historical dynamics.