Image
Multispectral Landsat 8 ESRI on-the-fly imagery superimposing the 2009-2017 difference in unvegetated to vegetated ratio calculated using classified NAIP imagery. Salt marsh along the Charleston, South Carolina coast relies on sediment import from the environment to remain stable. Areas in yellow are indicative of greater salt marsh change from vegetated to unvegetated. Change from vegetated to unvegetated indicate these areas may be vulnerable to open water conversion resulting in the loss of the salt marsh.   Keywords: UVVR, Jacob Stid, Adriana LeCompte, Derek Nguyen, Elspeth Gates​

Among the most productive ecosystems on earth, salt marshes provide crucial ecosystem services including water filtration, shoreline protection, storm surge buffering, and flood mitigation. Marshes are largely dependent on their sediment budget which can significantly vary across a region. Upstream land use change near Charleston, South Carolina, along with rising sea levels, are expected to alter sediment budgets and threaten marsh stability and long-term health. The unvegetated-vegetated ratio (UVVR) is a scalable and efficient method to assess vulnerability. This NASA DEVELOP project collaborated with the South Carolina Department of Natural Resources, the South Carolina Department of Health and Environmental Control, and the United States Geological Survey Woods Hole Coastal and Marine Science Center. Marsh vulnerability was analyzed using UVVR derived from Landsat 8 Operational Land Imager (OLI) and Landsat 7 Enhanced Thematic Mapper (ETM+) in conjunction with National Agriculture Imagery Program (NAIP) high-resolution aerial imagery. A Landsat random forest regression showed a low correlation (r2 = 0.247) between Landsat 7 ETM+ bands and NAIP aggregated UVVR suggesting the need for a more complex model and higher resolution sensors. Google Earth Engine scripting provided a novel approach to UVVR methodology that will allow decision makers to input new marsh areas and easily calculate UVVR without external data downloading.