Synthetic Aperture Radar (SAR) systems such as Sentinel 1 generate several million km2 of ocean imagery each month. Their analysis reveals many metoceanic phenomena, such as the presence of surfactants, sea ice, rain, convection zones or atmospheric fronts. If their detection is possible by an experienced analyst on a limited number of cases, the exploitation of all the SAR measurements remains for the moment impossible without the help of artificial intelligence (Colin et al., 2022, https://doi.org/10.3390/rs14040851). It is the role of this database to provide a set of examples of such detections to be used for their training. It is built as an extension of existing databases (Wang et al., 2019, https://doi.org/10.1002/gdj3.73) by two CLS engineers. Among the medium-term applications targeted are mapping of ocean surface states and a better understanding of air-sea interaction processes.