Oil Spill Monitoring in the North Sea Based on Deep Learning Using SAR Imagery

Kurzfassung

Oil pollution in the North Sea is mainly caused by shipping activities and offshore oil production. The North Sea contains busy international shipping lanes through the English Channel and to the Baltic Sea. In addition, hundreds of offshore oil and gas platforms are scattered throughout the North Sea. Oil leaks and discharges from offshore platforms, pipelines and ships result in surface films that threaten sensitive marine and coastal ecosystems such as the Wadden Sea. Continuous satellite-based monitoring is therefore needed to detect accidental and deliberate oil spills in this large area, coordinate their mitigation and identify possible polluters. In this contribution, we present an oil spill monitoring system for the North Sea based on Synthetic Aperture Radar (SAR) imagery from the Sentinel-1 satellite. For this, a diverse dataset with manually labeled SAR image patches including oil spills and typical other oceanic and atmospheric phenomena was generated. The labeling was supported by contextual information such as wind speed and the position of offshore platforms. This dataset was used to train a state-of-the-art deep learning based object detector. The detector was then validated with an additional dataset of labeled SAR images covering all seasons. Each detected oil spill is segmented to provide the actual extent of the polluted area. The oil spill monitoring system is designed for operational use. Therefore, the key metrics are the false negative and false discovery rates. By taking advantage of local image statistics, we show that the accuracy of the detector can be significantly improved. Starting from individual SAR scenes, the monitoring system automatically generates geocoded binary masks and relevant metadata of the detected oil spills. The method is optimized for speed and accuracy and can be adapted for upcoming SAR missions. With its potential for operational use, our oil spill monitoring system can serve as a basis for decision makers and early response to mitigate pollution.