Marsh collapse thresholds for coastal Louisiana estimated using elevation and vegetation index data

Forecasting marsh collapse in coastal Louisiana as a result of changes in sea-level rise, subsidence, and accretion deficits necessitates an understanding of thresholds beyond which inundation stress impedes marsh survival. The variability in thresholds at which different marsh types cease to occur (i.e., marsh collapse) is not well understood. We utilized remotely sensed imagery, field data, and elevation data to help gain insight into the relationships between vegetation health and inundation. A Normalized Difference Vegetation Index (NDVI) dataset was calculated using remotely sensed data at peak biomass (August) and used as a proxy for vegetation health and productivity. Statistics were calculated for NDVI values by marsh type for intermediate, brackish, and saline marsh in coastal Louisiana. Marsh-type specific NDVI values of 1.5 and 2 standard deviations below the mean were used as upper and lower limits to identify conditions indicative of collapse. As marshes seldom occur beyond these values, they are believed to represent a range within which marsh collapse is likely to occur. Inundation depth was selected as the primary candidate for evaluation of marsh collapse thresholds. Elevation relative to mean water level (MWL) was calculated by subtracting MWL from an elevation dataset compiled from multiple data types including light detection and ranging (lidar) and bathymetry. A polynomial cubic regression was used to examine a random subset of pixels to determine the relationship between elevation (relative to MWL) and NDVI. The marsh collapse uncertainty range values were found by locating the intercept of the regression line with the 1.5 and 2 standard deviations below the mean NDVI value for each marsh type. Results indicate marsh collapse uncertainty ranges of 30.7–35.8 cm below MWL for intermediate marsh, 20–25.6 cm below MWL for brackish marsh, and 16.9–23.5 cm below MWL for saline marsh. These values are thought to represent the ranges of inundation depths within which marsh collapse is probable.