Retrieval of atmospheric properties and surface bidirectional reflectances over land from POLDER/ADEOS

Abstract

Polarization and Directionality of the Earth's Reflectances (POLDER) is a new instrument devoted to the global observation of the polarization and directionality of solar radiation reflected by the Earth‐atmosphere system. It will fly onboard the ADEOS platform in 1996. This paper outlines the improvements expected from POLDER in the description of atmospheric aerosols and water vapor over land, and of surface bidirectional reflectances. It then gives a detailed description of the operational algorithms which are implemented in the “land surface and atmosphere over land” processing line. This line is part of an effort initiated by Centre National d'Etudes Spatiales (the French Space Agency) to develop lines of products in order to facilitate the exploration of POLDER's new capabilities by the international science community. Emphasis is given in this paper to the presentation of the principles, physical rationale, and elements of validation of the algorithms of this processing line. The main products are (1) for each orbit segment, the amount and type of aerosols, the water vapor content, and bidirectional reflectances corrected for atmospheric effects, and (2) every 10 days, global maps of surface directional signatures, of hemispherical surface reflectances, and of parameters describing the statistical distribution of aerosol and water vapor content. These products will be made available to all interested investigators. The most innovative algorithms of the processing line are (1) cloud detection, based on a series of tests involving reflectance thresholds, oxygen pressure estimates, and analysis of polarized radiance in the rainbow direction, (2) retrieval of aerosol optical thickness and type from directional polarized radiance measurements, and (3) retrieval of surface directional signature through an adjustment of a time series of directional reflectance measurements with a semiempirical bidirectional reflectance model.