Hydrocarbon micro-seepage detection from airborne hyper-spectral images by plant stress spectra based on the PROSPECT model

Abstract

Hydrocarbon micro-seepage can result in vegetation spectral anomalies. Early detection of spectral anomalies in plants stressed by hydrocarbon micro-seepage could help reveal oil and gas resources. In this study, the origin of plant spectral anomalies affected by hydrocarbon micro-seepage was measured using indoor simulation experiments. We analyzed wheat samples grown in a simulated hydrocarbon micro-seepage environment in a laboratory setting. The leaf mesophyll structure (N) values of plants in oil and gas micro-seepage regions were measured according to the content of measured biochemical parameters and spectra simulated by PROSPECT, a model for extracting hydrocarbon micro-seepage information from hyper-spectral images based on plant stress spectra. Spectral reflectance was simulated with N, chlorophyll content (Cab), water content (Cw) and dry matter content (Cm). Multivariate regression equations were established using varying gasoline volume as the dependent variable and spectral feature parameters exhibiting a high rate of change as the independent variables. We derived a regression equation with the highest correlation coefficient and applied it to airborne hyper-spectral data (CASI/SASI) in Qingyang Oilfield, where extracted information regarding hydrocarbon micro-seepage was matched with known oil-producing wells.