Detection of natural oil seeps signature from SST and ATI in South Yellow Sea combining ASTER and MODIS data

Abstract

Daytime Advanced Spaceborne Thermal Emission and reflection Radiometer (ASTER) data and the consequent nighttime Moderate Resolution Imaging Spectroradiometer (MODIS) data have been utilized to retrieve the sea surface temperature (SST) and apparent thermal inertia (ATI). They were used to detect the signature of natural oil seepage in Yingge Sea Basin (YSB) and South Yellow Sea (SYS). In this paper, ATI was used for the first time to detect the signature of natural oil seepage. ATI is an approximation of the real thermal inertia and can be expressed as a function of surface albedo and temperature difference between day and night. Surface albedo can be calculated by weighting spectral reflectivities in visible and near-infrared bands from ASTER data. The spectral reflectivities can be obtained after performing atmospheric correction using the 6S model. An iterative self-consistent split-window algorithm was employed to retrieve the daytime SSTs from ASTER data and nighttime SSTs from MODIS data. Because of the spatial resolution difference between ASTER and MODIS data, the nighttime SST derived from MODIS data was downscaled from 1 km to 90 m using the pixel block intensity modulation (PBIM) method. Two study areas, YSB and SYS, were selected to detect the possible signature of the natural oil seepage. The approximate SST and ATI values in the YSB where natural oil seepage actually developed were derived in advance. A look-up table of albedo, temperature difference and ATI for different sea waters with or without natural oil seepage was constructed. These were then used as prior knowledge to enclose the possible signatures of natural oil seepage in SYS. Our results show that there are some signatures of natural oil seepage in the study area in SYS. Compared with SST, ATI was more sensitive to the signature of natural oil seepage. The reason for this might be that the temperature difference was used in the calculation of the ATI, which decreased the influence of the SST accuracy on ATI values.

Acknowledgements

This work was supported by the MOST, China under Grant no. 2006AA12Z116. It is also supported by Beijing Natural Science Foundation under Grand No. 4102018, and by Beijing Municipal Organizing Department under Grant No. 070800905. It was also supported in part by Jurisdiction of Beijing Municipality under Grant Nos PHR200906138, PHR200907127 and PHR 20070101.