Automated Detection of Sargassum Invasions in the Caribbean Using Sentinel-1 Sar

Introductions of non-indigenous species such as seaweeds to new ecosystems are one of the major threats to biodiversity and resilience of coastal and marine habitats. Across the Atlantic into the Caribbean and Gulf of Mexico, recurring invasions of Sargassum seaweeds are currently monitored using satellite sensors such as OLCI, VIIRS and MODIS. These optical sensors are, however, limited by cloud, and their coarse spatial resolution prevents detection of seaweed patches smaller than approximately 2000 m² in size. Synthetic Aperture Radar (SAR) imagers, such as the sensor carried aboard Sentinel-1, may overcome both these limitations; detecting Sargassum through cloud at higher spatial resolutions.With a focus on open waters east of Barbados, we demonstrate the utility of Sentinel-1 GRDH SAR data for the detection of incoming Sargassum invasions. Application of a SARgassum index, which makes use of SAR data in predominantly VH-polarisation, improved contrast between Sargassum and background seawater. Downstream application of a Constant False Alarm Rate (CFAR) detector allowed for automated detection of floating Sargassum patches as small as 400 m² in size. This is a five-fold improvement on MODIS, proving the utility of Sentinel-1 for monitoring of Sargassum in coastal to open waters that may also be impacted by cloud. Though SAR data collected by Sentinel-1 is not currently included in existing Sargassum monitoring efforts over the Caribbean and Gulf of Mexico, our results suggest that it is a valuable but overlooked resource.