Computer Science
Scientific paper
Sep 2008
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2008epsc.conf..684d&link_type=abstract
European Planetary Science Congress 2008, Proceedings of the conference held 21-25 September, 2008 in Münster, Germany. Online a
Computer Science
Scientific paper
Abstract The Martian surface is mainly composed by an altered particulate mineral regolith, ranging from micrometric to bulk rocks sizes. A central goal in planetary science is determination of soils composition, in order to reconstruct the planets' evolutionary story at local and global scale. This goal is nowadays carried out by the support of remote sensing data analysis, coming from spacecraft carrying instruments suite able to collect radiation coming from the planet in a wide spectral range. A very useful spectral range to investigate soils mineral composition and atmospherics components dynamics is the thermal infrared (broadly between 1-50 μm) because primary spectral signatures of targets' components fall in this spectral range (i.e. stretching and roto-vibrational band of Si-O, O-H-O, C-O). The Planetary Fourier Spectrometer (PFS), onboard ESA's Mars Express spacecraft, collects radiation in the 1-45 μm range. High spectral resolution observations cover almost the whole Martian surface with a very large temporal basis (about 2 Martian years). The current PFS dataset is composed of over 500.000 spectral observations, enough for application of statistical analysis methods, so this study is carried out by mean of the Factor Analysis (FA) technique [1] that is demonstrated to be able to extract independently variable components and to recover the spectral end-member present in a spectral dataset. These data offer the chance to study minor mineral components of atmospheric particulate as well of the surface [2]. Surface analysis form remote planetary data needs the correction for atmospheric signature [2], including all the components that contribute to observed radiaton, as carbon dioxide and water vapour or atmospheric particulates (dust water ice). Planetary apparent emissivity spectra can be accurately modelled [1] by linear combination of atmospherics spectral shapes, when observation are taken in a region of low to moderate atmospheric opacity and with no distinctive surface features (high albedo regions). When this model does not accurately fit for the data, it usually needs an additional set of component to accurately reproduce the observation. This set is a suite (library) of pure mineral, intended to model the surface spectrum. Then, once the contribution of the atmospheric components is fixed, it is possible to extract the contribution due only to the soils from the observed radiation. To accomplish this procedure the exact shape of atmospherics components is needed. They are obtained by mean of FA technique from the PSF data, selected on a wide range of observational scenarios with varying atmospheric dust and water ice clouds opacities. The independently variable components in the dataset are extracted (Fig. 1), obtaining the spectral shape of those components and allowing the occasional monitoring of local and seasonal aerosol composition, morphology and temporal evolution. Our results show that derived atmospheric components are in agreement with previous TES results, showing a high degree of temporal uniformity in the mineral suspended haze (or at least of only one component of the dust), and allow to monitor the annual variation of the these atmospheric components, that is in again in good agreement with previous works [3]. References [1] Bandfield, J.L., Christensen, P.R., Smith, M.D. (2000) JGR, 105, 9573-9588. [2] Smith, M.D., Bandfield, J.L., Christensen, P.R. (2000) JGR, 105, 9589-9607. [3] Smith, M.D. (2004) Icarus ,167,148-165
D'Amore Mario
Helbert Jérôme
Maturilli Alessandro
Palomba Ernesto
Zinzi Angelo
No associations
LandOfFree
Principal Component Analysis and Target Transformation end-member recovery : application to last PFS MEX data. does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Principal Component Analysis and Target Transformation end-member recovery : application to last PFS MEX data., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Principal Component Analysis and Target Transformation end-member recovery : application to last PFS MEX data. will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1794119