The Heidelberg iterative cloud retrieval utilities (HICRU) and its application to GOME data
Résumé
Information about clouds, in particular the accurate identification of cloud free pixels, is crucial for the retrieval of tropospheric vertical column densities from space. The Heidelberg Iterative Cloud Retrieval Utilities (HICRU) retrieve effective cloud fraction using spectra of two instruments designed for trace gas retrievals from space: The Global Ozone Monitoring Experiment (GOME) on the European Remote Sensing Satellite (ERS-2) and the SCanning Imaging Absorption SpectroMeter for Atmospheric CHartographY (SCIAMACHY) on ENVISAT.
HICRU applies the widely used threshold method to the so-called Polarization Monitoring Devices (PMDs) with higher spatial resolution compared to the channels used for trace gas retrievals. Cloud retrieval and in particular the identification of cloud free pixels is improved by HICRU through a sophisticated, iterative retrieval of the thresholds which takes their dependency on different instrumental and geometrical parameters into account. The lower thresholds, which represent the surface albedo and strongly affect the results of the algorithm, are retrieved accurately through a four stage classification scheme using image sequence analysis.
The design and the results of the algorithm applied to GOME data are described and compared to several other cloud algorithms for GOME. The differences to other cloud algorithms are discussed with respect to the particular characteristics of the algorithms.
HICRU applies the widely used threshold method to the so-called Polarization Monitoring Devices (PMDs) with higher spatial resolution compared to the channels used for trace gas retrievals. Cloud retrieval and in particular the identification of cloud free pixels is improved by HICRU through a sophisticated, iterative retrieval of the thresholds which takes their dependency on different instrumental and geometrical parameters into account. The lower thresholds, which represent the surface albedo and strongly affect the results of the algorithm, are retrieved accurately through a four stage classification scheme using image sequence analysis.
The design and the results of the algorithm applied to GOME data are described and compared to several other cloud algorithms for GOME. The differences to other cloud algorithms are discussed with respect to the particular characteristics of the algorithms.
Domaines
Océan, Atmosphère
Origine : Accord explicite pour ce dépôt