Oil-spill forensics using two-dimensional gas chromatography: Differentiating highly correlated petroleum sources using peak manifold clusters

Abstract

Petroleum forensics for apportioning the environmental impact of oil spills necessitate quantitative differentiation between highly correlated biomarker distributions of neighboring oil sources. (GC × GC) generates high-resolution images that represent the complex hydrocarbon peak profiles of these petroleum biomarkers. As such, source differentiation reduces to the complex challenge of disambiguating the source-specific biomarker peak profile against strong regional commonalities, which are challenging to decorrelate using statistical techniques. We propose signal processing innovations that enhance recent methods in petroleum fingerprinting to achieve quantitative source differentiation. Specifically, we propose three related techniques: Peak topography maps; Peak Manifold clustering techniques; and Baseline interference mitigation.