Satellite-based detection of oil spill signature residual using synergy multispectral images of Landsat-8 OLI and Landsat-7 ETM+

Abstract

Satellite-based detection of oil spills is essential as it assists in reducing the risk of disasters occurring in the ocean environment to protect and reduce economic losses. The visible bands of Landsat-8 OLI and Landsat-7 ETM+ is a crucial means for detecting ocean oil spills that cause grave damage to the maritime ecosystem. Therefore, this study used the synergy of Landsat images to identify the residual oil spills along the Arabian Gulf region. Feature-based detection techniques by employing both region and point-based were adopted to compare with the multi-sensor's spatial-resolutions for detecting the oil spills. These multi-sensor data were used to extract the spill-pixels for signature's analyses. The variants against spill signatures were distinguished through calibration analysis. Results revealed that the pre-processed, as well as atmospherically corrected signatures of the spill, are reasonably agreed with the occurrence of spill records, demonstrating a high correlation (R2 > 0.85, p < 0.001). The calibration matrices have been formulated to calibrate spill pixels for Landsat-8 OLI and Landsat-7 ETM+, which serves as a crucial input for oil spill monitoring system against local uncertainties from look-alikes. Hence, this study will attribute in fast-tracking 2025 agenda of target 14.1 (reduce marine pollution) of sustainable development goal 14th set by United Nation.