Chemical, multispectral, and textural constraints on the composition and origin of rocks at the Mars Pathfinder landing site

Rocks at the Mars Pathfinder site are probably locally derived. Textures on rock surfaces may indicate volcanic, sedimentary, or impact‐generated rocks, but aeolian abration and dust coatings prevent unambiguous interpretation. Multispectral imaging has resolved four spectral classes of rocks: gray and red, which occur on different surfaces of the same rocks; pink, which is probably soil crusts; and maroon, which occurs as large boulders, mostly in the far field. Rocks are assigned to two spectral trends based on the position of peak reflectance: the primary spectral trend contains gray, red, and pink rocks; maroon rocks constitute the secondary spectral trend. The spatial pattern of spectral variations observed is oriented along the prevailing wind direction. The primary spectral trend arises from thin ferric coatings of aeolian dust on darker rocks. The secondary spectral trend is apparently due to coating by a different mineral, probably maghemite or ferrihydrite. A chronology based on rock spectra suggests that rounded maroon boulders constitute the oldest petrologic unit (a flood deposit), succeeded by smaller cobbles possibly deposited by impact, and followed by aeolian erosion and deposition. Nearly linear chemical trends in alpha proton X‐ray spectrometer rock compositions are interpreted as mixing lines between rock and adhering dust, a conclusion supported by a correlation between sulfur abundance and red/blue spectral ratio. Extrapolations of regression lines to zero sulfur give the composition of a presumed igneous rock. The chemistry and normative mineralogy of the sulfur‐free rock resemble common terrestrial volcanic rocks, and its classification corresponds to andesite. Igneous rocks of this composition may occur with clastic sedimentary rocks or impact melts and breccias. However, the spectral mottling expected on conglomerates or breccias is not observed in any APXS‐analyzed rocks. Interpretation of the rocks as andesites is complicated by absence of a “1 μm” pyroxene absorption band. Plausible explanations include impact glass, band masking by magnetite, or presence of calcium‐ and iron‐rich pyroxenes and olivine which push the absorption band minimum past the imager's spectral range. The inferred andesitic composition is most similar to terrestrial anorogenic icelandites, formed by fractionation of tholeiitic basaltic magmas. Early melting of a relatively primitive Martian mantle could produce an appropriate parent magma, supporting the