An Assessment of Surface Reflectance Obtained from Dark Target Algorithm Applied to Geostationary Satellite Measurements.

Abstract:

Separating surface contribution from measured top-of-atmosphere reflectance is a key process for retrieving aerosol properties from satellite remote sensing. Dark Target (DT) algorithm was designed to parameterize the land surface reflectance by using physical relationship between reflectance in visible (0.47, 0.66 μm) and shortwave-IR (2.12 μm) channel. Since the method does not require database accumulated over the previous years, it allows to the algorithm retrieves aerosol products from the beginning at launch. Therefore, the method has been widely applied to operating algorithms for global observation of geostationary orbit (GEO) and low Earth orbit satellites. A version of DT algorithm provides aerosol products from the Visible Infrared Imaging Radiometer Suite measurement, and a prototype DT algorithm has been applied for GEO measurements. The algorithms are successfully expanding the Moderate Resolution Imaging Spectroradiometer (MODIS) DT aerosol products in terms of time and spatial coverage. However, with regard to the differences in viewing geometry in sensors on different platform, issues have been raised with the application of current empirical relationship.

This study investigated the uncertainties associated with applying the current DT algorithm to GEO sensors, with a focus on parameterizing surface reflectance. The DT-surface reflectance obtained from the Advanced Himawari Imager and the Advanced Baseline Imager were compared to the atmospherically corrected reflectance, and the dependences of retrieval error on various factors were analyzed. The GEO surface reflectance showed good agreement with the atmospherically corrected surface reflectance. However, the error was relatively higher than the error in MODIS products, and it shows a dependency on scattering angle. A newly obtained spectral relationship has different sensitivity to the scattering angle from the current version also. In this regard, modification of the surface reflectance parameterization for GEO measurement was discussed here.