Estimating aboveground biomass in interior Alaska: statistical methods for coupling remotely sensed data with field observations to improve precision
Author
Babcock, Chad
Metadata
Show full item recordAbstract
The United State Forest Service’s (USFS) Forest Inventory and Analysis (FIA) program is charged by Congress to monitor the nation’s forestlands—survey growth, mortality and composition, among other aspects related to the status and trends of United States (US) forests. To accomplish this, the FIA repeatedly inventories a massive network of sample plots that spans most of the country’s forests. Although the FIA’s permanent forest inventory network is the largest in the world, it does not represent the entirety of the nation’s forested landscape. Interior Alaska contains approximately 15 percent of the nation’s total forestland but, until recently, has not been included in the FIA’s monitoring efforts. Being one of the most vulnerable ecosystems to climate change in the US, it is pivotal to begin monitoring the dynamics of interior Alaska’s forested landscape. The vastness of the remote Alaskan landscape makes thorough field-only inventories prohibitively expensive—we need new and innovative methods to track forest dynamics in Alaska. The USFS and NASA’s Carbon Monitoring System jointly funded a pilot project in 2013 titled Monitoring Forest Carbon Stocks in Interior Alaska to examine the potential of airborne and spaceborne remote sensing technologies to augment sparse collections of field data to obtain carbon estimates with acceptable levels of precision. This study will yield first-ever regional estimates of carbon stocks for interior Alaska’s Tanana Valley leveraging FIA inventory data—an important initial step in the process of developing a forest monitoring system for Alaska’s wilderness. The research summarized in the following dissertation examines several statistical methods for coupling field and remotely sensed information to estimate aboveground biomass and carbon stocks that can potentially be used for tracking forest carbon dynamics of interior Alaska. Both design- and model-based inferential paradigms are considered. This research is strongly focused on the appropriate characterization of uncertainty in the form of standard errors and confidence intervals. Results form this research will help decision makers meet the challenges of environmental change by providing statistically rigorous methodologies that can estimate interior Alaskan carbon stocks. Having reliable and transparent estimates of carbon and biomass stocks can inform carbon management choices about the role Alaskan forestlands play in the global carbon system.
Collections
- Forestry [391]