SPATIAL PATTERNS OF SAND COMPOSITION IN THE GRAN DESIERTO, MX DETERMINED FROM THERMAL INFRARED SPECTROSCOPY AND ASTER REMOTE SENSING DATA

The Gran Desierto in Sonora, Mexico is an important eolian depositional system located east of the Colorado River Valley, south of the Basin and Range Province and adjacent to the northern coast of the Gulf of California. The Sierra Pinacate volcanic complex, dominated by basalts, lies to the east of the main dune field. The center of the Gran Desierto overlies thick sequences of fluvial-deltaic sediments associated with the ancestral Colorado River. There are groups of dunes hypothesized to be genetically different eolian accumulations of different age and composition. Therefore, a number of possible local- to regional-scale sand transport pathways exist for sources of sand to the dunes.

Determination of the surface mineral composition was accomplished using thermal infrared (TIR) remote sensing data from the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) instrument. Because the mixing of emitted radiant energy is linearly related to the areal abundance of major dune forming minerals that have diagnostic spectral absorption features, the mineralogical composition was extracted using spectral linear deconvolution of the TIR emission spectra from both remote sensing and laboratory spectroscopy. TIR remote-sensing data were from a seamless, multispectral, radiometrically balanced image mosaic from the ASTER instrument. The laboratory analysis of eolian sand samples using high resolution thermal emission spectroscopy were used to validate remote sensing data.

The major spectral end-members of the samples were determined, reported here in decreasing order of areal abundance: quartz > plagioclase feldspar > potassium feldspar >> carbonate >> ferrohornblende. Spatial patterns of bulk mineral composition revealed the eolian sediment mixing dynamics and sediment transport pathways. The interpolated laboratory data provided hyperspectral resolution at specific sampling points, which compared well to the retrievals of sand composition from ASTER data. The majority of the sand in the Gran Desierto can be described by a two-source sediment mixing model, revealed by the spatial gradients between quartz-rich Colorado River source material and feldspar-rich local source material supplied by the Basin and Range and the Sonoyta River.