Sentinel 1 data for fuel spill detection. The case of the ships collision near Corsica

In the field of the maritime environment surveillance, the detection and mapping of oil on the water surface is one of the common use of the satellite images. The location of the spill should be obtained very precisely by means of image processing. It is nowadays standard practice to use Synthetic Aperture Radar (SAR) for producing this kind of thematic maps. In fact, the capability to acquire images in day-night and allweather conditions, covering large areas with frequent overpass, makes the active sensors the best for oil spill detection. In this study, we focused on the potential of the SAR data collected by the European Earth observation program Copernicus. Copernicus offers free radar satellite images with high spatial and temporal resolution. In particular, Sentinel-1 mission meets the features required by marine surveillance tools. Sentinel-1A and the Sentinel-1B satellites collect SAR images with a swath of 250 km in Interferometric Wide (IW) mode, covering wide area with high spatial resolution (10 m pixel size). The two-satellite constellation offers a repeat cycle of 6 days and a repeat frequency, considering ascending and descending orbit directions, of 1- 3 days in accordance with the latitude. The Sentinel-1 data are available for the Copernicus services within one hour of acquisition. In the present work, Sentinel-1 images have been processed to map the oil spill caused by the ships collision occurred near Corsica coasts in October 2018. For the SAR data classification, a near-real time procedure was developed through the Object-Based Image Analysis (OBIA).