Remote Estimation of Above Ground Forest Biomass Using LiDAR and Drone Imagery

As concerns around climate change continue to rise, accurate and efficient monitoring of our planet’s carbon stocks is becoming increasingly important. The carbon stocks of forests, one of the largest contributors to the Earth’s carbon cycle, have historically been monitored through measurements of Above Ground Biomass (AGB). These measurements are traditionally done by hand over long periods of manual labor; however, the recent rise of remote sensing methodology has provided another avenue for AGB estimation. A large amount of research has been conducted in the past decade regarding a wide range of methods for remote AGB estimation. Many of these methods, however, require large amounts of data to be gathered in order to be conducted in areas not recently analyzed. This study aims to test the feasibility of remote AGB estimation in an environment with low access to existing ground-truth data in the Appalachian area and develop reliable methods for AGB estimation in smaller-scale plots. These estimations are built off of existing equations relating tree height, canopy diameter, and genus directly to biomass. Through the use of drone imagery and machine learning, we are currently building a dataset that would allow rapid identification of tree genus by canopy on an image-by-image basis. Combining this with existing methods of tree measurement using public satellite LiDAR data, our goal is to develop both a benchmark dataset and trained machine learning model that could estimate tree biomass without the need for specialized equipment.