Remote-sensing-based maps of tidal marshes, both of their extents and carbon stocks, have the potential to play a key role in informing greenhouse gas inventories and implementing climate mitigation policies. Despite multiple efforts to map tidal marsh aboveground biomass regionally, differences in plant communities and site characteristics pose limitations for scaling to the national level. Given differences in plant functional type both within and across sites, the primary objective was to generate a single remote sensing model of tidal marsh aboveground biomass that 1) represents nationally diverse tidal marshes within the conterminous United States (CONUS), 2) is repeatable, consistent, and free or low cost, and 3) can generate aboveground carbon stock estimates when combined with data on percent carbon (C) in aboveground vegetation. To meet this objective, USGS scientists developed the first national-scale dataset of aboveground tidal marsh biomass, species composition, and aboveground plant %C from six CONUS regions: Cape Cod, MA; Chesapeake Bay, MD; Everglades, FL; Mississippi Delta, LA; San Francisco Bay, CA; and Puget Sound, WA.
Using the random forest algorithm, Sentinel-1 radar backscatter metrics and Landsat vegetation indices were tested as predictors of biomass. The final model, driven by six Landsat vegetation indices and with the soil-adjusted vegetation index as the most important (n = 409, RMSE = 310 g/m2, 10.3% normalized RMSE [root mean square error]), predicted biomass for a range of plant functional types defined by height, leaf angle, and growth form. Model error was reduced by scaling field-measured biomass by Landsat fraction green vegetation derived from object-based classification of National Agriculture Imagery Program imagery. Sentinel-1 radar backscatter metrics did not improve model performance but did demonstrate potential for distinguishing emergent from scrub-shrub marshes. The final model was used to generate 30-m-scale biomass maps for estuarine and palustrine emergent tidal marshes as indicated by a modified NOAA Coastal Change Analysis Program map for each region. With a mean plant %C of 44.1% (n = 1384, 95% C.I. = 43.99% - –44.37%), USGS scientists estimated mean aboveground carbon densities (Mg/Ha) and total carbon stocks for each wetland type for each region. Louisiana palustrine emergent marshes had the highest C density (2.67 +/-0.08 Mg/Ha) of all regions, while San Francisco Bay brackish/saline marshes had the highest C density of all estuarine emergent marshes (2.03 +/-0.06 Mg/Ha). This modeling and data synthesis effort will allow for aboveground C stocks in tidal marshes to be included for the first time in the 2018 U.S. Environmental Protection Agency Greenhouse Gas Inventory for coastal wetlands. As technical barriers have been reduced through the availability of free post-processed satellite data, cloud computing platforms, and open source software, this approach can potentially be applied globally as well.
Tidal marsh biomass maps of six CONUS regions based on a single Landsat-based random forest model. Moving from top, left to right: San Francisco Bay, CA; Cape Cod, MA; Everglades, FL; Nisqually National Wildlife Refuge, WA; Chesapeake Bay, MD; and Terrebonne and St. Mary Parishes, LA.