Estimating Forest metrics in interior Alaska with terrestrial Lidar.

Abstract

This thesis contains multiple research documents, including a report for the Forest Service, a detailed field guide, a metadata document, and a research proposal. Boreal forests make up a third of global forested area and play a vital role in global carbon storage, but their vulnerability to climate change necessitates improved methods for monitoring. We need accurate models of forest structure to understand how boreal forests will respond to climate change. Lidar remote sensing shows potential to fulfil this need by providing fine-scale models of individual trees. study assesses the application of Terrestrial Laser Scanning (TLS) to estimate Above Ground Biomass (AGB) in Alaska's boreal forests, focusing on its performance compared to traditional Forest Inventory and Analysis (FIA) methods. TLS, using the FARO Focus S360 scanner, detected significantly more trees than FIA, especially smaller trees that FIA does not measure due to its diameter threshold. These differences in stem count are most pronounced in black spruce and mixed forests. Trees from the segmented TLS point clouds were matched with FIA manually measured trees. The matched trees were used to train a Random Forest model to predict tree species. With predicted species for all TLS detected trees, a species specific allometric equation to predict biomass was employed on an individual tree level. Statistical tests, including Kruskal-Wallis and Dunn's post-hoc tests, revealed that forest type and species composition significantly influenced errors in TLS-derived metrics, with deciduous forests showing the lowest variability in biomass estimates. Due to its ability to capture smaller trees, TLS identifies a greater amount of biomass, particularly in black spruce forests. The findings highlight TLS's potential to improve biomass estimates but also emphasize the need for careful calibration and species-specific adjustments. Future work should focus on integrating TLS with airborne and mobile Lidar technologies to scale forest structure assessments across broader landscapes. The field guide in Appendix A. provides a framework for conducting terrestrial and mobile laser scans in forest inventory in Alaska's boreal forests. By standardizing data collection and analysis protocols, this field guide aims to support the U.S. Forest Service's Forest Inventory and Analysis (FIA) program and advance the use of Lidar for assessing forest structure, carbon dynamics, and resilience in boreal ecosystems. The metadata document in Appendix B. provides detailed descriptions of data collected during field campaigns in Alaskan. These datasets are the product of this study, containing details of tree structural metrics such as diameter at breast height (DBH), height, projected area, and alpha volume, calculated from TLS, as well as derived attributes like biomass, proximity, and species predictions. The research proposal in Appendix C., funded by the Gloria Barron Wilderness Society Scholarship, addresses the need for monitoring forest state changes in Alaskan wilderness areas. The study plans to integrate aerial laser scanning (ALS) and mobile laser scanning (MLS), to develop a scalable method for quantifying forest structural complexity and monitoring ecological trajectories in boreal forests.