In this work we have searched through Cassini/VIMS [1] hyperspectral cubes, selecting those data which have convenient viewing geometry and which overlap with Cassini/RADAR [2] footprints having comparable ground resolution, in order to properly look for correlations between the infrared and microwave ranges explored by the two instruments. In RADAR data we have considered two geophysical quantities: the normalized backscatter cross-section obtained from the scatterometer measurement, corrected for the incidence angle, and the brightness temperature determined from the radiometer measurement, as found in publicly available data products. In VIMS data, combining spatial and spectral information, we have selected some infrared wavelengths in the methane windows, which provide the best optical depth to measure surface reflectance. The two RADAR parameters are combined with the VIMS data, with estimated errors, to produce an aggregate data set, that we process using multivariate classification methods to identify homogeneous taxonomic units in the multivariate space of the samples. A first analysis has been done with the G-mode method [3], which has been successfully used in the past for the classification of such diverse data sets as lunar rock samples, asteroids and planetary surfaces. This method can be used without any a priori knowledge of the taxonomic structure of the observations, which is in fact provided by the classification. Furthermore, independence of variables and samples is not required, although the relationship between variables and samples needs to be known.
ANALYSIS OF SELECTED CASSINI VIMS AND RADAR DATA OVER THE SURFACE OF TITAN THROUGH MULTIVARIATE CLASSIFICATION METHODS
ML Moriconi;
2008
Abstract
In this work we have searched through Cassini/VIMS [1] hyperspectral cubes, selecting those data which have convenient viewing geometry and which overlap with Cassini/RADAR [2] footprints having comparable ground resolution, in order to properly look for correlations between the infrared and microwave ranges explored by the two instruments. In RADAR data we have considered two geophysical quantities: the normalized backscatter cross-section obtained from the scatterometer measurement, corrected for the incidence angle, and the brightness temperature determined from the radiometer measurement, as found in publicly available data products. In VIMS data, combining spatial and spectral information, we have selected some infrared wavelengths in the methane windows, which provide the best optical depth to measure surface reflectance. The two RADAR parameters are combined with the VIMS data, with estimated errors, to produce an aggregate data set, that we process using multivariate classification methods to identify homogeneous taxonomic units in the multivariate space of the samples. A first analysis has been done with the G-mode method [3], which has been successfully used in the past for the classification of such diverse data sets as lunar rock samples, asteroids and planetary surfaces. This method can be used without any a priori knowledge of the taxonomic structure of the observations, which is in fact provided by the classification. Furthermore, independence of variables and samples is not required, although the relationship between variables and samples needs to be known.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
