--> ABSTRACT: Offshore Hydrocarbon Seepage Mapping through Hyperspectral and Multispectral Data: A Possible Means for Remotely Estimating Oil API Gravity and Timing of Oil Exposition over Water

Datapages, Inc.Print this page

Offshore Hydrocarbon Seepage Mapping through Hyperspectral and Multispectral Data: A Possible Means for Remotely Estimating Oil API Gravity and Timing of Oil Exposition over Water

Lammoglia, Talita 1; Souza Filho, Carlos R.1
(1) Department of Geology and Natural Resources, Geosciences Institute, University of Campinas, Campinas, Brazil.

This work aims to evaluate optical spectral reflectance properties of crude oil, oil films over water and oil-water emulsions and the use of their spectral identity to classify RS imagery. Chemometric techniques were tested to estimate the API gravity and the age of water-oil emulsions based on their temporal laboratory spectral responses.Laboratory measurements of 17 oil samples indicated that spectra of different crude oils, oil films over water and oil-water emulsions allow the indirect determination of their API gravity and viscosity and, subordinately, the percentage of their SARA components. Also, spectra of pure crude oils and crude oil films on water, re-sampled both to the 357 and 9 VNIR-SWIR channels of the PropecTIR and ASTER sensors, respectively, allowed their qualitative separation based on intrinsic spectral signatures. The chemometric models proved efficient in predicting oil status considering full-resolution lab spectra (2150 bands-VNIR-SWIR) and spectra resampled to equivalent hyperspectral (ProspecTIR) and multispectral (ASTER) sensor’s resolutions. An oil seepage recorded by ASTER in November/2004 at the Brazilian coast was approached to test the predictive models. The oil seep is located over the Marlim Sul oil field (Campos Basin), which has a petroleum of 17-24 API gravity. The remotely predicted API gravity through ASTER’s retrieved spectra and PLS processing was 19.6 +/-1,37. Considering that oil began to seep by 18/November/2004 and that the ASTER scene was recorded by 20/November/2004, traces of the oil seepage that are most distant from its source should have around 5 days of environmental exposition. The remotely prediction using ASTER data indicated between 3-4 days of oil exposition on the ocean, confirming that the model can successfully forecast the timeframe of oil exposition in the ocean (i.e., the relative “age” of the seepage). The limited spectral resolution of the ASTER sensor implies in less accurate and less robust estimates, which can be highly enhanced by hyperspectral sensors. The spectral libraries and method proposed in this study can be reproduced to other oceanic areas in order to remotely approximate the API gravity and age of noticeable natural oil seepages or anthropogenic oil spills. This type of information is optimal for seepage or leakage vectoring. This research demonstrates the outstanding potential of spectroscopy, chemometric analysis and ASTER data for low-cost, offshore hydrocarbon exploration.

 

AAPG Search and Discovery Article #90135©2011 AAPG International Conference and Exhibition, Milan, Italy, 23-26 October 2011.