An Approach of Vicarious Calibration of Sentinel-2 Satellite Multispectral Image Based on Spectral Library for Mapping Oil Spills

Abstract

Abstract. Sentinel-2 satellite Multispectral Image (MSI) is one of the recent advancement of satellite optical imaging for detecting and tracking oil spills. MSI equipped with enhanced radiometric and spatial resolutions, apart from relatively high temporal resolution of every 5 days revisit capability. Both systematic errors of the geometric and radiometric of level 1 and 2 data were successfully treated before any data download for users’ levels applications. As such, leaving the random errors, crucially to be minimized to enable oil spill detection and tracking due to non-discernible absolute signatures of spills against the scene background and the look-alikes. The magnitude of these random errors’ minimization and the efficacy of the MSI absolute signatures within visible bands for oil spills is very crucial. However, it is rarely reported; in fact, it is a new issue to be addressed accordingly. The calibrating tool was created with oil spill spots revealed by the official authorities. Whereas, the spill pixels are identified in the corresponding pre-processed Sentinel MSI image using region growing segmentation algorithm. These spill pixels grown were analyzed against the RGB bands, logistically regressed against the oil spill via a spectral library of the crude oil type. Originated from Arabian Gulf region with an average film thickness of 0.5 to 4 mm; reporting a calibrating function in a form gain and bias corrections for RGB bands, respectively. The results indicated that calibrated MSI spill pixels have higher correlation (r2 > 0.85, p < 0.001). As the signature variations were used to formulate calibration matrices for spills identified from satellite images which can be used for processing of spill monitoring system.

Sensors, Features, and Machine Learning for Oil Spill Detection and Monitoring: A Review

“…The radiometric correction of optical images is required to mitigate the atmospheric effects to improve the identification of oil spills [53,66] and remove the radiometric sensor aging effects and radiometric discrepancy among sensors (i.e., Landsat TM and ETM+) [166]. Atmospheric correction softens the atmospheric effects by eliminating the influence of the atmospheric molecules and aerosol scattering [71] and improving the extraction of real surface parameters from satellite images (i.e., surface reflectance, emissivity, and surface temperature) [167]. This correction is considered in different oil spill studies [30,32,47,49,53,62,64,168].…”

Monitoring oil spill in Norilsk, Russia using satellite data

“…Green and red light bands of remotely sensed data were used as optimal bands for detecting the oil spills of sea surface 3,25,49,70,80 . Different algorithms were used to correct aberrations, color, and light levels and to process images 39,40 . In this study, the occurrence of the oil spill is studied by developing a true color composite and decorrelated image, and the occurrence of snow and ice, water, vegetation, and wetlands are studied and monitored by developing false-color composites and band ratios indices images 3,49,54,81 .…”

“…The finest spectral resolution of hyperspectral remote sensing at the nanometer (nm) level was also utilized to identify spectral characteristics of oil spills [34][35][36][37][38] . Research studies showed the potential use of Sentinel-2 data to detect and map the oil spills spread over the gulf or sea 30,39,40 .…”

Assessing LAPAN-A3 Satellite with Line Imager Space Application (LISA) Sensor for Oil Spill Detection

“…In other research, Arslan [16] assessed oil spills using Landsat-8 multispectral sensors in Fener (Ufak) Island, Cesme, Turkey. Kolokoussis and Karathanassi [17] and Althawadi and Hashim [18] mapped oil spills using Sentinel-2 MSI. Determination of band combination is one commonly used method for detecting oil spills.…”