Remote Sensing of Vegetation and Aquatic Parameters for Modeling Coastal Marsh Response to Sea Level Rise

Kristin Byrd, USGS, kbyrd@usgs.gov (Presenter)
Lisamarie Windham-Myers, United States Geological Survey, lwindham@usgs.gov
Lisa Schile, Smithsonian, lmschile@gmail.com
Thomas Leeuw, Sequoia Scientific, tleeuw@sequoiasci.com
Emmanuel Boss, University of Maine, emmanuel.boss@maine.edu
Michael Vasey, SF Bay National Estuarine Research Reserve, mvasey@sfsu.edu
Matthew Ferner, SF Bay National Estuarine Research Reserve, mferner@sfsu.edu

The NOAA National Estuarine Research Reserve System (NERRS) and surrounding communities need a verifiable, spatially explicit forecasting model of tidal marsh response to sea level rise (SLR) to address the potential impacts of SLR on coastal ecosystems and dependent wildlife species. The Marsh Equilibrium Model (MEM) is a one-dimensional mechanistic elevation-based soil cohort model that models marsh elevation change based on feedbacks between field-measured organic (plant biomass) and inorganic (suspended sediment) inputs. Working at Rush Ranch, a San Francisco Bay NERR site, we tested the feasibility of obtaining two important MEM inputs, peak biomass and annual average suspended sediment concentration (SSC), from Landsat 8. We tested the sensitivity of MEM to remotely sensed inputs as compared to field measured inputs, and to error associated with the remote sensing inputs. We produced a biomass map of Rush Ranch that applied the Wide Dynamic Range Vegetation Index (WDRVI) (ρNIR*0.2 – ρR)/(ρNIR*0.2+ρR) to fully vegetated pixels and the simple ratio index (ρRed/ρGreen) to pixels with a mixed signal of vegetation and water. RMSE for top 90th percentile biomass values was 326 g/m2. We also produced a time series of SSC with a single band semi-analytical model based on local mass specific absorbing and scattering properties (R2 = 0.66, RMSE = 3.38 mg/L). Comparison of Landsat 8 and field-based MEM inputs found no significant difference in projections across 95% of the marsh plain area at 100 years, with both projections illustrating a subtle “sinking” of the marsh. Integration of remote sensing data would transform MEM into a spatial model for forecasting coastal marsh vegetation distributions and habitat suitability for special status species to aid regional decision making.

Presentation: 2015_Poster_Byrd_131_68.pdf (1336k)

Presentation Type:  Poster

Session:  Theme 1: Tracking habitat change through new integrative approaches and products   (Mon 1:30 PM)

Associated Project(s): 

• Byrd, Kristin: Forecasting Coastal Habitat Distributions through Fusion of Earth Observations, Process Models and Citizen Science: A Climate Change Adaptation Tool for the NOAA National Estuarine Research Reserve System ...details

Poster Location ID: 131