Over the course of the last three decades, Synthetic Aperture Radar (SAR) has proven itself to be an effective monitoring technology for marine applications. The clear benefits of using SAR as opposed to optical devices is that SAR is insensitive to cloud cover, lighting conditions and can also provide imagery to a high degree of resolution. Given these benefits, there is a large incentive to implement SAR as a primary detection mechanism for marine oil spills due to the fact that SAR is capable of reliably providing data on a semi-daily basis. With increasing levels of maritime traffic due to declines in Arctic multiyear sea ice as well as risks associated with oil and gas exploration in the Arctic, being able to derive important geophysical information on the state of an oil slick is important for the decision-making process of first responders and clean-up personnel. This thesis is concerned with attempting to determine the dielectric properties of oil slick using SAR. The dielectric constant is a proxy for the volumetric water/oil content within an oil slick. This is due to the fact that when pure crude oil is inserted to the marine environment, it becomes subjected to a host of processes collectively referred to as weathering. Throughout these processes, oil-in-water emulsions can form that alter the dielectric properties of an oil slick resulting in a substance that has a dielectric value between that of pure crude oil and pure sea water depending on the volume of sea water present within an emulsion. In this thesis, we first apply a two-scale theoretical backscattering model to quad-polarimetric Radarsat-2 data of verified oil slick acquired during oil-on-water exercises conducted in the North Sea between the years of 2011-2013, acquired under varying wind conditions and incidence angles. The results showed realistic values for the dielectric constant given auxiliary information on the state of the slicks. However, no in-situ information was available to verify the model. A unique set of data was then acquired during the NORSE2019 oil-on-water experiment by DLRs F-SAR instrument in full quad-polarimetric X-, S- and L-bands. This data set was used to verify the model approach used in this thesis as well as to investigate the time variability of the discharged slicks using a stability measure in conjunction with a novel polarimetric feature that exploits the multifrequency aspect of the data set. The work presented in this thesis sheds light on the on-going discussion on the use of SAR for marine slick characterization.
Paper I: Quigley, C., Brekke, C. & Eltoft, T. (2020). Retrieval of Marine Surface Slick Dielectric Properties from Radarsat-2 Data via a Polarimetric Two-Scale Model. IEEE Transactions on Geoscience and Remote Sensing, 58(7), 5162-5178. Published version not available in Munin due to publisher’s restrictions.
Published version available at https://doi.org/10.1109/TGRS.2020.2973724
Accepted manuscript version available in Munin at https://hdl.handle.net/10037/20939.
Paper II: Quigley, C., Brekke, C. & Eltoft, T. (2020). Comparison Between Dielectric Inversion Results from Synthetic Aperture Radar Co- and Quad-Polarimetric Data via a Polarimetric Two-Scale Model. IEEE Transactions on Geoscience and Remote Sensing, Early Access. Published version not available in Munin due to publisher’s restrictions.
Published version available at https://doi.org/10.1109/TGRS.2020.3038366
Accepted manuscript version available in Munin at https://hdl.handle.net/10037/20846.
Paper III: Quigley, C., Brekke, C. & Eltoft, T. Analysis of the Dielectric Retrieval Capabilities of Simultaneous X- S- and L-band Airborne Synthetic Aperture Radar for Oil Spill Monitoring Applications. (Submitted manuscript).