HR: 0800h
AN: P51B-0494    [Abstracts]
TI: Mineralogy of the Moon by multiple endmember spectral unmixing of Clementine UVVIS and NIR data
AU: * Combe, J
EM: jcombe@spacescience.org
AF: Bear Fight Center, 22 Fiddler's Road P.O. Box 667, Winthrop, WA 98862, United States
AU: Kramer, G
EM: gkramer@spacescience.org
AF: Bear Fight Center, 22 Fiddler's Road P.O. Box 667, Winthrop, WA 98862, United States
AU: Thomas, M B
EM: mccordtb@aol.com
AF: Bear Fight Center, 22 Fiddler's Road P.O. Box 667, Winthrop, WA 98862, United States
AB: Global observations of the Moon have been made by the imaging spectrometers onboard the Clementine spacecraft in 1994 [1]. These data have 5 channels in the Ultra-Violet Visible (UVVIS) in the range 450-1000 nm and 6 in the Near Infra-Red (NIR) in the range 1100-2780 nm. UVVIS global mosaic has been widely used in spectral analysis (e.g. [2, 3]). Calibration issues in the NIR data did not favor spectral studies using the full wavelength range [4]. Only a few analyses have been made in the NIR using different calibration approaches (e.g. [5]), despite potential valuable information to discriminate more minerals. A new calibration has finally been applied on the NIR [6, 7] and a mosaic calibrated in Bidirectional Reflectance (BRDF) has been released [8]. Quantitative mapping of minerals is possible using Clementine data 1) by converting BRDF into single scattering albedo [9, 10] and 2) by performing spectral linear unmixing [11]. Multiple-Endmember Spectral Mixture Analysis (MESMA, [12]) has been successfully applied with UVVIS data [11]. This method allows spectral unmixing to be calculated using a limited number of components, even if a lot more spectral endmembers are available in the reference spectral library. We have analyzed newly calibrated Clementine UVVIS and NIR data using a multiple-endmember linear spectral linear unmixing algorithm [13]. The model works either with spectra collected from the image or from laboratory measurements of pure minerals. Components are selected on spectral interpretation and knowledge of lunar soils and surface processes. Variation in the spectral slope varies with surface maturity and TiO2 content [14]. Including a slope endmember in the reference spectral library allows the mineralogy to be mapped independently of maturity effects. [1] Nozette S., et. al., 1994, Science, 266 [2] Lucey P. G., 2004, GRL 31. [3] Pieters C. M. et al., 2006, Icarus, 184. [4] Lucey P. G. et al., 1998, LPSC abstract 1576. [5] Le Mouelic et al., 1999, JGR 104. [6] Eliason E. M. et al., 2003, LPSC abstract 2093. [7] Cahill et al., 2004, LPSC abstract 1469. [8] Gaddis et al., 2007, PDS Volumes USA_NASA_PDS_CL_5001 through 5078. [9] Hapke B., 1981, JGR 86. [10] Johnson P. E., 1983, JGR 88. [11] Li L. and Mustard J. F., 2003, JGR 108. [12] Roberts D. A. et al., 1998, RSE 65. [13] Combe J.-Ph. et al., submitted to PSS. [14] Charette M. P. et al., 1974, JGR 79.
DE: 1026 Composition of the moon
DE: 3934 Optical, infrared, and Raman spectroscopy
DE: 5464 Remote sensing
DE: 5470 Surface materials and properties
DE: 6250 Moon (1221)
SC: Planetary Sciences [P]
MN: 2007 Fall Meeting