The Surface Composition of Martian Low Albedo Regions Revisited

Introduction Low albedo regions on Mars are often interpreted as outcrops of volcanic rocks. Mineral models of the thermal emission spectra obtained by TES indicate that the martian dark regions are characterized by basaltic surface material: large fraction of feldspar and one high-calcium pyroxene [1]. The data from the IR spectrometer ISM onboard Phobos-2 show that the composition of these layers is rich in pyroxenes and contains a significant signature of hydratation [2]. A systematic comparison of TES and ISM data suggests that variations in the vis-NIR observations could be controlled by dust or other thermally neutral materials [3]. New high resolution visible images from the Mars Observer Camera and IR thermal images of the THEMIS spectrometer onboard Mars Odyssey show that the low albedo regions are correlated with dark sand dunes, sand sheets and eolian mantling [e.g, 4]. This implies that the dark material may not always correspond to in-situ rock outcrops. Even if it is important to remember that the different observational techniques (visible, NIR and thermal) are sensitive to different characteristics of the martian surface, the understanding of discrepancies of the compositional analysis from different measurements and the nature of low albedo layers is essential 1to understand their erosional history, and 2to interpret the IR data of future spectrometers like OMEGA and PFS onboard Mars Express. The purpose of this work is to revisit the surface composition of dark regions by modeling ISM spectra representative of dark regions with a radiative transfer theory. Spectral data The data used are from the PDS archive on www ias fr cdp Base ISM INDEX HTM . Two ISM windows (Aurorae and Syrtis-Isidis) were used in this study. The basic approach is to extract the spectra with albedo (with aerosols scattering and photometric corrections) lower than 15%. This selection (about 700 spectra for each window) should cover most of the terrains studied previously with the MGM method [2]. The spectra are characterized by 1and 2-micron absorptions and gray/slightly red slope between 0.8 and 2.5 microns. Choice of the scattering model We choose to use the Shkuratov radiative transfer theory for fitting the spectra. This geometrical optics model based on the slab approximation for calculating the albedo of a particle has been compared to other scattering models [6] and tested with laboratory mineral mixtures [7]. Choice of the optical constants and end-members We select the surface composition of each spectrum by trying to satisfy the following spectral characteristics: low albedo, shape and depth of 1and 2-micron absorptions and spectral slope. Lowand high-calcium pyroxenes were obviously included in the scattering calculations. Spectrally featureless low albedo component in near-infrared to lower the average spectral reflectance is also required. Oxides such as magnetite display this neutral opaque behavior. Hematite (a Figure 1: Upper: Three synthetic spectra of an intimate mixture of one obsidian, two pyroxenes and a dark component (here magnetite). The size of coarse particles for each component is indicated. The concentration of obsidian is 20%, 5% and 0%. Lower: Same except for obsidian replaced by amphibole.