Determining Crystal Abundance in Glassy Lavas: Combining Laboratory Infrared Spectroscopy With Remote Sensing

C. Eisinger and J. Fink, Arizona State University, Department of Geological Sciences Box 871404, Tempe, AZ 85287

M. Ramsey, Department of Geology and Planetary Science, 200 SRCC Building, University of Pittsburgh, Pittsburgh, PA 15260

Variations in the surface composition of high silica glassy lavas can provide information about the emplacement history, eruptive style, and structure of flows and domes. However, traditional field mapping of these variations is commonly time consuming, incomplete and impossible for planetary and remote terrestrial environments. Thermal infrared (TIR) remote sensing can provide a more efficient and effective method for mapping this chemical variability. Both high spectral resolution laboratory data and high spatial resolution airborne thermal images were used to evaluate the abundance and composition of crystals in silicic glasses.

Image and laboratory thermal emission spectra were collected from the compositionally mixed Glass Mountain flow in northern California, which has a bulk silica content from 57% to 75%. The multispectral image data, acquired using the MODIS/ASTER (MASTER) airborne simulator, were reduced to emissivity and compared to laboratory emission spectra of 40 samples. Data were modeled using a linear deconvolution algorithm to calculate compositional abundances for two end-member suites dacite bulk rocks and the other containing common phenocryst minerals. Crystal contents derived from the deconvolution approach of the laboratory spectra ranged from 0% in glassy obsidians to more than 48% in the most mafic samples. These modeled abundances generally agreed with values derived from a petrographic analysis, but were higher on average. One possible explanation for this bias is an increased emitted infrared signal due to the presence of microlites and nanolites.

Application of the deconvolution-model to the MASTER remote sensing data produced crystal abundance maps that highlight zones of heterogeneity. However, complications due to instrument calibration, the effects of vesicularity, and the moderate spectral resolution limit the accuracy of this approach. As techniques are improved, spectral data from spaceborne instruments, such as ASTER and THEMIS, can be used to characterize glassy lavas on Earth and other planets. Laboratory spectroscopy provides a baseline for this work.

------------------
Submitted: American Geophysical Union Fall Meeting
Date: December 6 - 10, 2002