Abstract:

New York City’s coastal wetlands “centered on Jamaica Bay’s tidal marshes” provide storm buffering, carbon storage, water filtration, and habitat but have been degraded by relative sea-level rise, sediment alteration, eutrophication, and urbanization. We test a reproducible, multi-sensor workflow to map vegetation and track dynamics with two pipelines:(1) PlanetScope surface-reflectance imagery stacked with a USGS DEM and classified in R using a trained Random Forest to produce class and confidence GeoTIFFs; and (2) Sentinel-2 composites in Google Earth Engine generating seasonal NDVI (2016″2024) and annual NDWI (2016″2024) with robust cloud/cirrus masking. A Chesapeake Bay benchmark produced accurate five-class maps. In Jamaica Bay, the model reliably separated open water from low marsh (Spartina alterniflora) but under-represented higher-elevation and edge communities, indicating domain-shift and feature-set limits. NDVI showed strong seasonality (summer peaks, winter minimal) and interannual variability consistent with restoration gains and edge erosion; NDWI captured dynamic wetness, including expanding/contracting ponds and wave-washed fringes. The approach supports post-Sandy management by delivering repeatable indicators of marsh extent, condition, and hydrologic state. It highlights priorities for Jamaica Bay” specific retraining, probability-aware mapping, expanded predictors (texture, tidal frequency, LiDAR), and spatial cross-validation. The results are from the NOAA EPP/MSI CSC NERTO graduate internship project conducted with NOAA mentor Dr. Veronica Lance of the CoastWatch/OceanWatch/PolarWatch Program, National Environmental Satellite, Data, and Information Service (NESDIS). The NERTO aligns with NOAA CSC CESSRST-II’s mission to advance earth system science, remote sensing, and data-driven environmental solutions, in support of NOAA’s goals of a Weather-Ready Nation, Resilient Coastal Communities, and Climate Adaptation. The NERTO, Leveraging Satellite Earth Observations to Understand Wetland Ecosystem Services for Coastal Resilience, strengthened the intern’s skills by applying a multi-sensor workflow to map vegetation and track marsh dynamics in Jamaica Bay using PlanetScope classification in R andSentinel-2 NDVI/NDWI time series in Google Earth Engine. The project improved understanding of wetland change drivers and NOAA research practices while enhancing scientific communication, technical reporting, and collaborative analysis for coastal resilience.