Offshore oil seepage visible from space : a Synthetic Aperture Radar (SAR) based automatic detection, mapping and quantification system
|Other Titles:||Marine Ölaustritte vom Weltraum aus gesehen : ein automatische System zur Detektierung, Kartierung und Quantifizierung, basierend auf Radar mit Synthetischer Apertur (SAR)||Authors:||Suresh, Gopika||Supervisor:||Bohrmann, Gerhard||1. Expert:||Bohrmann, Gerhard||2. Expert:||Notholt, Justus||Abstract:||
Offshore oil seepage is believed to be the largest source of marine oil, yet very few of their locations and seepage fluxes have been discovered and reported. Natural oil seep sites are important as they serve as potential energy sources and because they are hosts to a very varied marine ecosystem. These seeps can also be associated with gas hydrates and methane emissions and hence, locating natural oil seeps can provide locations where the sources of greenhouse gases could be studied and quantified. A quantification of the amount of crude oil released from natural oil seeps is important as it can be used to set a background against which the excess anthropogenic sources of marine oil can be checked. This will provide an estimate of the 'contamination' of marine waters from anthropogenic sources. Until the onset of remote sensing techniques, field measurements and techniques like hydroacoustic measurements or piston core analysis were used to obtain knowledge about the geological settings of the seeps. The remote sensing techniques either involved manual or semi-automatic image analysis. An automatic algorithm that could quantitatively and qualitatively estimate the locations of oil seeps around the world would reduce the time and costs involved by a considerable margin. Synthetic Aperture Radar (SAR) sensors provide an illumination and weather independent source of ocean images that can be used to detect offshore oil seeps. Oil slicks on the ocean surface dampen the small wind driven waves present on the ocean surface and appear darker against the brighter ocean surface. They can, hence, be detected in SAR image. With the launch of the latest Sentinel-1 satellite aimed at providing free SAR data, an algorithm that detects oil slicks and estimates seep location is very beneficial. The global data coverage and the reduction of processing times for the large amounts of SAR data would be unmatchable. The aim of this thesis was to create such an algorithm that could automatically detect oil slicks in SAR images, map the location of the estimated oil seeps and quantify their seepage fluxes. The thesis consists of three studies that are compiled into one of more manuscripts that are published, accepted for publication or ready for submission. The first study of this thesis involves the creation of the Automatic Seep Location Estimator (ASLE) which detects oil slicks in marine SAR images and estimates offshore oil seepage sites. This, the first fully automatic oil seep location estimation algorithm, has been implemented in the programming language Python and has been tested and validated on ENVISAT images of the Black Sea. The second study reported in this thesis focuses on the optimisation of the created ASLE and comparison of the ASLE with other existing algorithms. It also describes the efficiency of the ASLE with respect to other existing algorithms and the results show that the ASLE can successfully detect seeps of active seepages. The third study aimed to provide the status of the offshore seepage in the southern Gulf of Mexico estimated from the ASLE using SAR images from ENVISAT and RADARSAT-1. The ASLE was used to detect natural oil slicks from SAR images and estimate the locations of feeding seeps. The estimated seep locations and the slicks contributing to these estimations were then analysed to quantify their seepage fluxes and rates. The three case studies illustrate that an automatic offshore seepage detection and estimation system such as the Automatic Seep Location Estimator (ASLE) is very beneficial in order to locate global oil seeps and estimate global seepage fluxes. It provides a technique to detect offshore seeps and their seepage fluxes in a fast and highly efficient manner by using Synthetic Aperture Radar images. This allows global accessibility of offshore oil seepage sites. The availability of large amounts of historic SAR datasets, the presence of 5 active SAR satellites and the latest launch of the European Space Agency satellite Sentinel-1, which provides free data, shows that there is no shortage in the availability of SAR data. The result of the work done in this thesis provides a means to utilise this large SAR dataset for the purpose of offshore oil seepage detection and offshore seepage related geophysical applications. The created system will be an important tool in the future not just to estimate offshore seepage in local seas but in global oceans that are otherwise challenging for field analysis.
|Keywords:||oil seeps, natural oil slicks, Synthetic Aperture Radar, SAR, automatic detection, object classification, ENVISAT, RADARSAT, ASLE||Issue Date:||20-Apr-2015||URN:||urn:nbn:de:gbv:46-00104459-12||Institution:||Universität Bremen||Faculty:||FB5 Geowissenschaften|
|Appears in Collections:||Dissertationen|
checked on Sep 28, 2020
checked on Sep 28, 2020
Items in Media are protected by copyright, with all rights reserved, unless otherwise indicated.