Blue Carbon

Scaling-up Carbon Fluxes Using Remote Sensing, Machine Learning, and Regional Climate Models for Projecting Productivity of Salt Marshes

The primary goal is to accurately estimate the dynamics of carbon (C) exchange in flooded eastern US coastal salt marshes (LTER sites) and understand how flooding variability influences the dynamics of C assimilation. Three selected study sites have in-situ flux tower data available from AMERIFLUXL 1) the Plum Island Ecosystems LTER in northeastern Massachusetts (US-PLM) 2) the Virginia Coast Reserve LTER (US-VCR) and 3) the Georgia Coastal Ecosystem LTER (US-GCE). This data will be used to train and validate the Machine Learning approaches to model MODIS/VIIRS GPP. Thus, we will combine spectral estimates of species-specific Light Use Efficiency (LUE) and fPAR to form the basis of a VIIRS GPP algorithm.

The focus of our upcoming research will be on predicting future shifts in C dynamics in wetlands. Moreover, accurate mapping of C fluxes allows for the spatial prediction of future GPP fluxes by taking into account the impact of flooding on C dynamics and the opposite as well. Understanding the role of flooding and regional climate in regulating salt marsh GPP will provide new insights into blue carbon source-sink dynamics under future climate conditions.

Funding and Collaborators:

• University of Georgia

• NSF (Georgia Coastal Ecosystems LTER OCE1237140 and OCE1832178)