Atmospheric Correction of Envisat/Meris Data over Case Ii Waters: the Use of Black Pixel Assumption in Oxygen and Water Vapour Absorption Bands

Jingjing Zhu,Jin Chen,Bunkei Matsushita,Wei Yang,Takehiko Fukushima

DOI: https://doi.org/10.1080/01431161.2011.632656

IF: 3.531

2012-01-01

International Journal of Remote Sensing

Abstract:The Medium Resolution Imaging Spectrometer MERIS sensor, with its good physical design, can provide excellent data for water colour monitoring. However, owing to the shortage of shortwave-infrared SWIR bands, the traditional near-infrared NIR–SWIR algorithm for atmospheric correction in inland turbid case II waters cannot be extended to the MERIS data directly, which limits its applications. In this study, we developed a modified NIR black pixel method for atmospheric correction of MERIS data in inland turbid case II waters. In the new method, two special NIR bands provided by MERIS data, an oxygen absorption band O2 A-band, 761 nm and a water vapour absorption band vapour A-band, 900 nm, were introduced to keep the assumption of zero water-leaving reflectance valid according to the fact that both atmospheric transmittance and water-leaving reflectance are very small at these two bands. After addressing the aerosol wavelength dependence for the cases of single-and multiple-scattering conditions, we further validated the new method in two case lakes Lake Dianchi in China and Lake Kasumigaura in Japan by comparing the results with in situ measurements and other atmospheric correction algorithms, including Self-Contained Atmospheric Parameters Estimation for MERIS data SCAPE-M and the Basic ERS European Remote Sensing Satellite & ENVISAT Environmental Satellite AATSR Advanced Along-Track Scanning Radiometer and MERIS BEAM processor. We found that the proposed method had acceptable accuracy in the bands within 560–754 nm MERIS bands 5–10 average absolute deviation AAD = 0.0081, average deviation AD = 0.0074, which are commonly used in the estimation models of chlorophyll-a chl-a concentrations. In addition, the performance of the new method was superior to that of the BEAM processor and only slightly worse than that of SCAPE-M in these bands. Considering its acceptable accuracy and simplicity both in principle and at implementation compared with the SCAPE-M method, the new method provides an option for atmospheric correction of MERIS data in inland turbid case II waters with applications aiming for chl-a estimation.