GSA Annual Meeting in Phoenix, Arizona, USA - 2019

Paper No. 295-8
Presentation Time: 3:25 PM

HYPERSPECTRAL ANALYSIS OF MINERAL MIXTURES IN ALLUVIAL FAN SEDIMENTS USING THE USGS MICA ALGORITHM: FROM LABORATORY TO FIELD TO IMAGING SPECTROMETER SCALES OF DETECTION (Invited Presentation)


HUBBARD, Bernard E., US Geological Survey, 12201 Sunrise Valley Drive, Reston, VA 20192, MARS, John C., U.S. Geological Survey, 12201 Sunrise Valley Dr., MS 954, Reston, VA 20192, HOOPER, Donald, WEX Foundation, 110 E. Houston Street, 7th Floor, San Antonio, TX 78205 and KOKALY, Raymond F., U.S. Geological Survey, MS964 Box 25046 DFC, Denver, CO 80225

Airborne Visible InfraRed Imaging Spectrometer (AVIRIS) data were flown across the Salton Sea area of Southern California between 2013 and 2018 in preparation for the HyspIRI mission (now the SBG – Surface Biology Geology mission). The AVIRIS data were calibrated to reflectance and validated using ground truth field spectral measurements acquired under natural sunlight and atmospheric conditions. Field measurements were acquired along 34 traverses, ranging in length from 21 to 353 m, across various alluvial fans (AF) on the flanks of the Salton Sea. Samples from each field site were collected for spectral lab analysis using an artificial light source. The alluvial fan surfaces included well-developed desert pavements and varnish coatings in upper (or proximal) fan areas, to clastic gravel sieve and channel deposits in medial and distal areas (“bar-and-swale” features). Swale deposits were mostly coarse sand, but occasionally finer sands, clay-rich soil crusts, and biotic shells were collected in distal fan areas. Mixing influences from sediments deposited by ancestral pluvial Lake Cahuilla are most evident in the distal parts of the fans towards the modern Salton Sea.

The USGS Material Identification and Characterization Algorithm (MICA) was used to identify and map the most abundant minerals dominating each lab, field, and AVIRIS spectrum. MICA utilizes continuum removal and linear regression procedures to compare observed spectral absorption features to those diagnostic of reference materials in a spectral library. MICA was used to identify and map clay, mica, carbonate, and iron-bearing minerals that were described in more detail previously (Hubbard et al., 2015 – GSA abst. w. prog., v. 47, no. 7, p. 74). Our MICA results show reasonable agreement between corresponding minerals identified in the field, laboratory, and AVIRIS data representing the same areas of each AF. Notably, the best matches occur in AFs derived from the Chocolate and Orocopia Mountains to the north and east (rather than those to the west) of the Salton Sea. This is perhaps due to smaller grain-size differences between bar-and-swale materials, more uniform desert pavement development, and locations above the ancient shoreline at those sites.