Proceedings of the 2022 Mississippi Water Resources Conference

Year: 2022 Authors: Wu W., Grimes E., Suir G.

To improve understanding on the efficacy of large-scale freshwater and sediment diversion projects in restoring coastal wetlands, we aim to evaluate the long-term changes in wetland vegetation in diversion outfall areas.

We selected two diversion outfall areas (Caernarvon and Davis Pond, LA, U.S.) as the study sites and compared them to the reference wetlands nearby with similar pre-diversion vegetation compositions. We implemented multi-level Bayesian models to evaluate 1) how peak-season Landsat-derived Normalized Difference Vegetation Index (NDVI) changed over time in reference to the initiations of the freshwater diversions, and 2) how peak-season NDVI responded to mean and variability of inundation and salinity, key abiotic factors that would be affected by freshwater diversions.

Analysis showed differences in temporal trends of NDVI for the Caernarvon project, and differences in how inundation and salinity impacted NDVI between the reference and diversion sites, indicating that diversions played a key role in affecting vegetation. For the Caernarvon diversion project, variability of water level (inundation) and salinity and mean salinity were important to affect NDVI, while mean water level and mean salinity were important for the Davis Pond diversion project. The variability has not been traditionally accounted for in quantitative models but can be more important to consider than the mean levels.

The study provides a modeling framework that can comprehensively evaluate freshwater diversions with uncertainties accounted for. It will inform future use of freshwater diversions for wetland restoration.