SEEING THE CARBON FOR THE TREES - USING REMOTE SENSING TO MODEL FOREST CARBON

Like many mid-sized cities across the United States, Chattanooga, TN is experiencing rapid urban growth. As metropolitan areas like Chattanooga continue to expand, more of Earth’s unique forest ecosystems are destroyed to make way for urban development. In general, the majority of vegetation present within, and oftentimes surrounding, a proposed urban development site will be effectively removed, resulting in an overall loss of biodiversity. Once extracted from a landscape, vegetation will naturally biodegrade, releasing virtually all the carbon sequestered in its biomass during its life cycle back to the atmosphere as carbon dioxide. In order to preserve the future health and longevity of Chattanooga’s existing biomass and carbon pool within trees, this research aimed to generate a model using digital image processing, remote sensing, and field measurements to estimate above ground biomass. Low cost methods with minimal time investment were prioritized to better understand the potential carbon sequestered in the biomass of urban trees, with the ultimate goal being to support administrative decision-making processes relative to urban development, regional forest health, and biodiversity enhancement. The campus of the University of Tennessee at Chattanooga (UTC), which accurately represents a standard city block in Chattanooga, was used as the preliminary study site for this research. The multispectral images acquired by both Sentinel-2 and PlanetScope satellites were used in this study. From these images, a number of vegetation indices, including the Normalized Difference Vegetation Index (NDVI), the Enhanced Vegetation Index (EVI), and the Modified Soil Adjusted Vegetation Index (MSAVI) were generated and used as predictor variables in a series of simple linear regression analyses. Results from this pilot study found a significant relationship between MSAVI and field-measured biomass values (R² = .7276). Future research will explore multiple linear regression, utilizing both structural predictor variables, such as canopy height, and spectral predictor variables, such as vegetation indices, to model sequestered carbon across larger areas, such as Hamilton County, TN.