Quantitative characterization of marine oil slick by satellite near-infrared imagery and oil drift modelling: the Fun Shai Hai case study

Abstract

Near-infrared (NIR) satellite images of the oil spill event caused by the Fu Shan Hai wreck on 31 May 2003 in the waters between Sweden and Denmark were compared with numerical simulations provided by the MIKE 21 oil drift model. Assuming a skewed probability density function (pdf) of oil parcel thicknesses, a model of the NIR image oil–water contrast reflectance was developed to characterize the expected oil slick distribution in terms of average and maximum oil slick thickness. Since MIKE 21 Spill Analysis (SA) also allows non-uniform distribution of oil volume within the oil slick, both distributions were thus compared by coincidence of the Moderate Resolution Imaging Spectroradiometer (MODIS/Aqua) acquisition, which imaged the oil slick 3 days after the oil spill started. Results showed an excellent agreement in the numerical values of both the expected average and the maximum thickness. In addition, repartition of the oil volume within the slick in the usual thin (sheen) and thick (brown) parts resulted, consistent with the empirical rule of 20% and 80% of the total oil volume, respectively.

Acknowledgement

This work was supported by the project, Flight Risks Mitigation and Nowcasting at Airports, funded by Apulia Region, POFESR 2007–2013. MODIS images were freely downloaded from the NASA archive available at http://ladsweb.nascom.nasa.gov/index.html. Atmospheric correction of the MODIS image used for this work was performed with the SeaDAS 6.0 software package, which was freely downloaded from http://oceancolor.gsfc.nasa.gov/seadas/.