GRSG 34th Conference 2023

Title: Oil spill characterisation with SAR

Author: Priscila Martins Oliveira da Costa


Oil spill characterization is a resilient topic in the literature. Economic and environmental impacts arising from oil occurrence in maritime zones foster the continuity of research on this theme. Petroleum floating on the sea surface can originate from ships, oil platforms, refineries, and natural seepages. Remote Sensing is used extensively in oil spill monitoring since it allows for collecting data and generating information in large areas. Among the sensors, the Synthetic Aperture Radar (SAR) stands out, especially because its operability is independent of solar radiation and clouds. Petroleum floating on the sea surface becomes visible in SAR images due to the dampening of short gravity-capillary waves that are responsible for the radar backscattering, which makes the oil appear as darker patches in the seawater surface in the image.

Despite advances made in the characterization of oil with SAR data, there is still much to advance regarding the discrimination between petroleum and targets of similar response, the response of oil in different polarization and SAR bands, effects of weathering on oil backscatter, as well as physical characterization with radar images. SAR sensors, previously on board only on airborne and satellite platforms, have gained a new imaging perspective with the increasing development of unmanned aerial vehicles (UAV). UAV-borne SAR systems can assist in oil spill response providing high-resolution local data, generating information at a lower cost and in less time than conventional platforms. In this context, this research intends to contribute to the classification and characterization of petroleum in SAR images through data acquired by a UAV-borne SAR system. To achieve this goal, a field experiment was designed to simulate oil leaks on a real scale.

The experiment included two tanks filled with simulated seawater where five and thirty liters of crude oil were leaked. During the experiment, the system was exposed to weathering processes uninterruptedly. The leaks were imaged in an unprecedented way, using an innovative UAV-borne SAR system that operates with three microwave frequencies simultaneously, including C, L and P bands, and four channels (two interferometric C band antennas (VV polarization), one P band antenna (HH polarization), and one L band antenna (HH, VV, VH and HV polarization). The system flew regularly over the experiment area for seventy-nine days between August and October 2022, acquiring data on programmed dates.

The research is still in progress, but preliminary results showed that the SAR system was able to identify differences between oil and clean saltwater backscattering responses in both tanks. The backscattering behavior of VV polarization from bands C and L, corresponding to days 1, 29 and 79 of the exposed oil, showed that clean saltwater displayed higher backscattering values than oil. Along the temporal series, a variation of backscattering values was also observed, which can be indicative of thickness variation, as well as weathering processes of the oil layers, which will be investigated. These preliminary results show the potential of UAV-borne multi-frequency SAR systems to qualify oil in the ocean.