THERMAL INFRARED-EMISSION SPECTROSCOPY OF NATURAL SURFACES - APPLICATION TO DESERT VARNISH COATINGS ON ROCKS

Citation
Pr. Christensen et St. Harrison, THERMAL INFRARED-EMISSION SPECTROSCOPY OF NATURAL SURFACES - APPLICATION TO DESERT VARNISH COATINGS ON ROCKS, J GEO R-SOL, 98(B11), 1993, pp. 19819-19834
Citations number
47
Categorie Soggetti
Geosciences, Interdisciplinary
Journal title
JOURNAL OF GEOPHYSICAL RESEARCH-SOLID EARTH
ISSN journal
21699313 → ACNP
Volume
98
Issue
B11
Year of publication
1993
Pages
19819 - 19834
Database
ISI
SICI code
2169-9313(1993)98:B11<19819:TISONS>2.0.ZU;2-T
Abstract
Thermal infrared spectroscopy has become an increasingly important too l for remote compositional analysis and geologic mapping. Most publish ed laboratory measurements have been obtained in bidirectional reflect ion or transmission, whereas remotely sensed thermal infrared data are obtained by measuring the emitted energy. Section 2 of this paper des cribes a laboratory technique for determining calibrated emissivities of natural surfaces. Equations are developed to account for the energy reflected from the environment and to determine directly the sample t emperature from measurements of hot and cold blackbody targets. Two me thods for determining emissivity are developed: one in which only a ho t sample measurement is made and the reflected background energy is re moved by modeling, and a second in which the sample is cooled and the reflected energy is measured directly. Relative emissivity can be obta ined to approximately 1% and absolute emissivities can be obtained to 2-15%, depending on the validity of the assumption that the emissivity of the sample is unity at some wavelength. The emission data agree we ll with the hemispherically integrated reflection data but point out p roblems associated with bidirectional reflectance measurements. Sectio n 3 applies emissivity measurements to the study of layered surfaces c onsisting of desert varnish coatings on granite and granodiorite rock suites. Two linear models are developed: the first assumes linear mixi ng of independent emission from the substrate and varnish (checkerboar d model); the second models transmission through an absorbing/emitting medium. Regardless of whether the varnish is or is not relatively tra nsparent and strongly absorptive, the spectral effect of varnish incre ases linearly with varnish thickness, indicating that thick patches of varnish dominate the spectral properties. As a result, median varnish thickness can be determined from spectral measurements. In addition, the composition of a substrate can be estimated through varnish layers up to 40-50 mum in median thickness, and the composition of the varni sh material can be determined if the substrate material is known. The varnishes studied are composed primarily of clay materials, consistent with previous studies of varnish composition.