Team profile by members of the Quantifying Wetland Carbon Emissions in the Southeastern United States team

Wetland ecosystems are highly productive and biologically diverse. While covering only 6% of the Earth’s land surface, wetlands provide a myriad of ecosystem services, such as storing one-third of the organic carbon on the planet, enhancing water quality, mitigating erosion and regulating stream flow.

Unfortunately, climate change – including global warming and sea-level rise – poses significant challenges to their functioning. One potential consequence is the acceleration of carbon emissions, which could undermine wetlands’ ability to sequester carbon efficiently, potentially shifting them from carbon sinks to carbon sources.

This study focuses on the southeastern United States, which is home to nearly half of the wetlands in the conterminous U.S. but is a region that has studied less than its northern counterpart. Our objectives are twofold: 1) to identify the key climatological variables influencing wetland CO₂ fluxes, and 2) to quantify the spatio-temporal patterns of wetland CO₂ fluxes across the entire southeastern U.S., thereby evaluating the potential impacts of climate change on wetlands.

To achieve these objectives, we first trained several machine learning (ML) models (including random forest (RF), support vector machine, and artificial neural network) using measured CO₂ flux data from AmeriFlux to find the optimal relationships between wetland CO₂ fluxes and environmental variables. The RF model demonstrated superior performance and was subsequently used to extrapolate historical wetland CO₂ fluxes in the southeastern U.S., using spatial climate data from the NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP-CMIP6).

Our findings highlighted the critical role of incoming shortwave radiation (SW_IN) and the Palmer Drought Severity Index (PDSI) as key environmental predictors of wetland CO₂ fluxes. Wetland CO₂ fluxes decrease with the incoming shortwave radiation until reaching a point of light saturation (SW_IN > 300 W m^-2). Additionally, the response of wetland CO₂ fluxes to PDSI indicates that both drought (PDSI < -2) and extreme flooding (PDSI > 4) conditions result in near-zero or positive CO₂ fluxes (wetlands emit CO₂ into the atmosphere).

The high-spatial-resolution (0.25° × 0.25°) and long-term (1950-2014) monthly gridded regional wetland CO₂ flux product over the southeastern U.S. developed in this study provides valuable insights. Notably, our estimations showed that wetlands along the mid-Atlantic coast and the Mississippi Delta served as significant CO₂ sinks, whereas the Everglades emerged as notable CO₂ sources. This product can inform targeted and effective wetland conservation and restoration efforts and serve as a benchmark and additional constraint for future wetland carbon flux modeling and upscaling studies in the study region.

Sunlight and Flooding/Drought Conditions Dominate Carbon Fluxes in Wetland Across the Southeastern United States

Poster by Aicha Slaitane, Divyansh Jain, Qiuying Liao, Zizhuo Chen, Keqi He and Wenhong Li