Sentinel-2 image transformation methods for mapping oil spill - A case study with Wakashio oil spill in the Indian Ocean, off Mauritius

MethodsX. 2021 Mar 27:8:101327. doi: 10.1016/j.mex.2021.101327. eCollection 2021.

Abstract

Although several indices have been constructed and available at the Index database (IDB) for Sentinel-2 satellite to map and study several earth resources, no indices have been developed to map oil spill. We constructed band ratios (5 + 6)/7, (3 + 4)/2, (11+12)/8 and 3/2, (3 + 4)/2, (6 + 7)/5 using the high-resolution MSI (multi-spectral instrument) visible-near infrared-shortwave infrared spectral bands of Sentinel-2 by summing-up the bands representing the shoulders of absorption features as numerator and the band located nearest to the absorption feature as denominator to discriminate oil spill, and demonstrate the potential of this method to map the Wakashio oil spill which occurred in the Indian Ocean, off Mauritius. The resulted images discriminated the oil spill well. We also decorrelated the spectral bands 4, 3 and 2 by studying the spectral band absorptions and discriminated the spill as very thick, thick and thin. The results of decorrelation stretch method exhibited the distribution of types of oil spill in a different tone, distinctly. Both the image transformation methods (band ratios and decorrelation stretch methods) showed their capability to map oil spills, and these methods are recommended to use for similar spectral bands of other sensors to map oil spills.•This study demonstrated the application of band ratios and decorrelation stretch methods to map oil spill.•The methods discriminated the oil spill off Mauritius, and showed spill thicknesses from the Sentinel-2 data.•The new methods are recommended to use for the spectral bands of other sensors to map oil spill.

Keywords: Decorrelation stretch; Image processing; Indices; Oil spill thickness; Sentinel-2.