Conference proceeding
Fingerprinting the Refugio oil spill using topographic signal processing of two-dimensional gas chromatographic images
OCEANS 2017 - Anchorage, pp.1-4
OCEANS (Anchorage, Alaska, USA, 09/18/2017 - 09/21/2017)
09/2017
Abstract
We examine petroleum forensics from a pattern recognition and feature separation perspective in this work. Apportioning the environmental impact of oil spills is important for marine pollution studies. Robust fingerprinting of an unknown sample from a petroleum-rich locale remains a data science challenge. Crude petroleum is a complex mixture, and as such, the fingerprint of a petroleum source can be discovered as the signature profile of hydrocarbon peaks corresponding to the biomarker compounds, which are well-known for their recalcitrance to environmental weathering. In this work, we apply recently proposed peak topography mapping techniques to examine the GCxGC topography of the Archean region based on four representative crude oil samples collected from the locale of the Refugio spill. Specifically, we compare the robustness of match between samples from the leaking pipeline of the Refugio spill against the match between oil samples from other local sources.
Details
- Title: Subtitle
- Fingerprinting the Refugio oil spill using topographic signal processing of two-dimensional gas chromatographic images
- Creators
- Rachel Bruflodt - University of IowaRobert K Nelson - Dept. of Marine Chem. & Geochem., Woods Hole Oceanogr. Instn., Woods Hole, MA, USAEleanor C Arrington - University of California, Santa BarbaraDavid Valentine - University of California, Santa BarbaraAnanya Sen Gupta - University of IowaKarin Lemkau - University of California, Santa BarbaraVeronika Kivenson - University of California, Santa BarbaraChristopher M Reddy - Dept. of Marine Chem. & Geochem., Woods Hole Oceanogr. Instn., Woods Hole, MA, USA
- Resource Type
- Conference proceeding
- Publication Details
- OCEANS 2017 - Anchorage, pp.1-4
- Conference
- OCEANS (Anchorage, Alaska, USA, 09/18/2017 - 09/21/2017)
- Publisher
- IEEE
- Language
- English
- Date published
- 09/2017
- Academic Unit
- Electrical and Computer Engineering
- Record Identifier
- 9984198001402771
Metrics
1 Record Views