From Data Gaps to Forest Futures: Mapping Current Conditions and Estimating Carbon Vulnerabilities in Southeast Alaska and Coastal British Columbia
Loading...
Date
2025-02-24
Authors
Lamping, James
Journal Title
Journal ISSN
Volume Title
Publisher
University of Oregon
Abstract
The temperate rainforests of Southeast Alaska and Coastal British Columbia are globally significant for their role in carbon storage and cycling, hosting some of the highest aboveground carbon densities in the world and large reserves of late-seral forest. Recent shifts in forest management practices have transitioned the region towards the conservation of late-seral forest, with the Tongass National Forest implementing young-growth management and British Columbia setting large areas aside for conservation. However, this region lacks estimates of forest composition and structure that are spatially complete across political boundaries. There is also uncertainty in how the interaction among young-growth management, climate change, and wind, the main driver of natural disturbance, will affect the future carbon storage capacities of these forests. The research presented is divided into three primary chapters. The first chapter utilizes a GNN modeling approach, integrating regional forest plot data with environmental predictors to estimate aboveground biomass, species-level biomass, basal area, and three other structural attributes across a vast and ecologically diverse landscape. The study identifies key environmental variables influencing forest structure and provides comprehensive maps highlighting spatial patterns of forest attributes. The second chapter focuses on the conservation of old-growth and mature forests, employing the GNN method to classify forest seral stages. It examines the distribution and extent of old growth and mature forests, contextualizing them within various protected areas and assessing their vulnerability to management policies. The third chapter models future carbon and species vulnerabilities under various scenarios of management and climate change using the LANDIS-II landscape model. Potential shifts in forest composition and carbon storage are quantified.
Description
Keywords
carbon, Forest structure, Gradient Nearest Neighbor, LANDIS-II, landscape model, old growth