Coastal wetland vegetation assemblages in Coastal Louisiana (1984–2025)

Plant communities, including those found in coastal wetlands, often reflect ecosystem processes and conditions at local and landscape scales more effectively than any other set of factors, and different communities can influence processes such as hydrodynamics, vertical accretion, and carbon burial in distinct ways. Thus, accurate land cover data in the region is fundamental for formulating sound planning of restoration efforts. Though considerable effort has been put forth to monitor the zonation of marsh vegetation communities in coastal Louisiana for over 30 years, the spatial and temporal resolution of these monitoring endeavors are often insufficient to adequately support restoration planning. To address this gap, we used machine learning (random forests; RF) and cloud computing (Google Earth Engine) to develop a new Landsat-based, land-cover dataset. The geospatial dataset depicts wetland vegetation community types defined in a previous study at annual (1985-2025) time steps at 30-m resolution.