P31A-01
What do we know of Mercury's composition?
Little is known of Mercury's bulk composition and the range of petrologic diversity within its crust and mantle. Our best constraint on the planet's bulk composition is an inference from its high-uncompressed density - Mercury has the highest metal to silicate ratio of all the terrestrial planets. However we do not know if this ratio is original or due to a catastrophic redistribution of materials early in its history. At a smaller scale Mariner 10 and Earth based telescopic data indicate that Mercury's crust may have compositional and mineralogic affinities with the lunar crust. The same data allow the tenuous interpretation that Mercury's mantle contains relatively low ferrous iron. Intriguing observations of Mercury's surface bound exosphere show a relatively high concentration of sodium, a finding not easily reconciled with other evidence that Mercury is the most refractory planet. This paper will review our knowledge of Mercury's composition and preview the MESSENGER mission's first flyby of Mercury that will occur in January of 2008.
P31A-02 INVITED
Lunar Surface Composition and Diversity
From the exploration of the Moon carried out by the Apollo missions and subsequent analysis of samples returned from the Moon, much is known of the Moon's composition and the range of rocks and minerals that constitute its crust. Rock compositions, mineral chemistry, and isotopic characteristics discriminate crustal rocks into (1) the ferroan-anorthositic suite (early, primary crust, complementary to the mafic mantle that formed from solidification of a magma ocean and that produced basaltic volcanism later in Moon's history), (2) the magnesian suite, which appears to be intrusive into the early crust, and (3) the alkali suite, which includes more evolved chemical differentiates of magmatic processes. The magnesian and alkali suites may be related to each other by extended magmatic fractionation, but the spatial relationship of these rock types within the lunar crust is not well known. Global remote sensing done by Clementine and Lunar Prospector in the 1990s showed that surface expressions of crustal composition vary strongly and broadly across the lunar surface. The northern far-side highlands were shown to be highly anorthositic, and a large region of the western nearside was shown to be strongly enriched in thorium and other heat-producing and incompatible elements. Perhaps as a consequence, the latter was the locus of extensive volcanic activity on the Moon, although the degree to which the surface expression reflects the character of the crust at depth and the underlying mantle is not known. Another key unknown is the composition of the lower crust. The enormous South Pole-Aitken basin on the Moon's southern far side provides a window into this lower crust. Remote sensing shows it to be mafic (Fe and Mg-rich) and to exhibit a modest thorium enrichment; however, the thorium content is significantly less than that of the near-side Procellarum KREEP Terrane, suggesting that the deep crust is not everywhere similarly enriched in heat- producing elements. The inferred global distribution of materials therefore reflects a strong global asymmetry that may relate to the very early differentiation of the Moon. That the Procellarum region exhibits subdued topography and includes a wide range of volcanic materials indicates that the compositional character of the crust there is coupled with the underlying mantle, consistent with an early differentiation process that sequestered much of the radiogenic heat production into this part of the Moon. Scientifically rich targets for future exploration to better understand the crust and its diversity include the far-side South Pole-Aitken basin and some of the more extensively fractionated volcanic materials of the Procellarum KREEP Terrane such as occur in the Aristarchus region and in the extensive lava flows of Western Procellarum.
P31A-03
Using Lunar Regolith Composition as a Geologic Tool
Despite that lunar surface rocks have been broken up, pulverized, and mixed to depths of 10's to 100's of meters by meteoroid impacts since the crust formed, the surface retains, both locally and regionally, a record of the geologic diversity of the crust. Mixing caused by the distribution of impact ejecta can be far reaching; witness the rays of Tycho. The process has nevertheless been inefficient. Compositional data obtained from orbit show a 10x range in iron concentration, and an even greater range in concentrations of incompatible elements, over the lunar surface. At all Apollo sites, any regolith sample contains a large variety of compositionally diverse rock types. Regolith compositions vary, e.g., from 80:20 to 20:80 across geologic boundaries (e.g., mare:highlands) on the scale of Apollo traverse distances (up to 10 km) because the relative proportions of the constituent rocks types change with lateral distance. Mature soils (long surface exposure) are rarely dominated by one lithology although immature soils (fresh ejecta) may be. Some Apollo regolith cores show distinct compositional variation with depth over decimeters because vertical mixing is inefficient. For all but the most volatile of lithophile elements mass-balance is preserved during impact melting and mixing (unlike many terrestrial processes that involve fluids). Thus, bulk compositions of lunar soils and breccias can be modeled as linear mixtures of the compositions of the target rocks or, inversely, constituent rock types can be inferred by mathematically unmixing regolith compositions, given what is known about lunar lithologies from sample studies. Over the past 25 years, 40-45 meteorites have been found on Earth that were blasted off the Moon from random locations by asteroidal impacts. Most of the meteorites are regolith or fragmental breccias from near the lunar surface. By comparing compositions of 100's of small (2-4 mm) lithic fragments from Apollo regoliths with those of the lunar meteorites, we can show that the Apollo collection is unrepresentative in containing a significant proportion of Th-rich rocks from the anomalous Procellarum KREEP Terrane and that Apollo rocks and regolith reflect large-scale mixing between the Procellarum KREEP Terrane and the Feldspathic Highlands Terrane as a result of the Imbrium and Serenitatis basin-forming impacts.
P31A-04 INVITED
Compositions of Igneous Rocks on Venus
The limited available data on Venus' igneous rocks (from landed geochemistry and radar-based geomorphology and elevation data) are all consistent with basalt of one variety or another. Venus' shield volcanos have shallow slopes, comparable to terrestrial basalt shields; Venus' lava plains are consistent with extensive flows of fluid lava, i.e. basalt; the margins of distinct lava flows are consistent also with fluid like basalt. A few small volcanic constructs (e.g., pancake domes) could represent more silicic lavas, but could also represent basalt extrusives, either crystal-rich or with very slow effusion rates. The Venera and VEGA landers, technical and scientific triumphs that they were, provided limited and imprecise constraints on the chemistry of Venus' basalts. Their landing sites were all in the lowland plains, and so did not sample rocks from any highland: shield volcanos, tesserae, nor the unique plateau and high mountains of Ishtar Terra. The V/V analyses for Mg, Al, and Mn are little more than 2σ detections, and V/V returned no data on Na, Cr, Ni, P and other minor/trace elements. The V/V analyses for K, U, and Th (by γ rays) are imprecise, except for one (Venera 8) with extremely high K (~4% K2O) and one (Venera 9) with a super- chondritic Th/U abundance ratio (at the 2σ level). Even with this imprecise limited data, a few inferences are fairly sound. [1] The FeO content of Venus basalts is similar to those of Earth basalts, suggesting a comparable mantle composition and thus a similar-sized core. [2] The range of K abundances suggests significant processes of depletion and enrichment in incompatible elements, consistent with repeated or extensive igneous processing. [3] The super-chondritic Th/U value measured by V9 is difficult to generate in low- pressure silicate melt/crystal fractionations. This Th/U value could possibly represent garnet fractionation in the mantle source, action of an ionic fluid (like carbonate melt) or even action of liquid water. [4] The Venus basalts, as a whole, have sub-chondritic Ca/Al. This relative deficit in Ca could arise from weathering at Venus's surface, or could be a primary characteristic possibly reflecting an origin as melted eclogite.
P31A-05
Geochemistry of Asteroid 4 Vesta: HED Meteorites as Constraints for Interpreting Gamma- Ray and Neutron Spectra from the Dawn Mission
Rotational spectra of Vesta have revealed compositionally distinct units that are smaller than the footprint of the GRaND instrument on the Dawn orbiting spacecraft. Geochemical analyses by GRaND will thus represent lithologic mixtures. HED meteorites, thought to have been derived from Vesta, can be used to quantify the admixed components. Using selected meteorites representing basaltic eucrite, cumulate eucrite, and diogenite as endmembers, we have modeled the chemical compositions of other HED meteorites. Element ratios are more accurately determined by this method than are absolute element abundances. A mixing model utilizing molar (Mg+Fe)/Si versus Al/Si ratios accurately calculates all major and minor element abundances. Surface analyses of Vesta can thus be deconvolved into the relative proportions of these HED lithologies. In addition, the model calculates the abundances of elements not analyzed by GRaND, such as Na, Mn, and Cr. The full set of elements enables the calculation of normative mineralogy, which can be compared with minerals identified by Dawn's VIR spectrometer. Another model, substituting Ca/Si for Al/Si, allows the direct estimation of olivine proportions in olivine-bearing diogenites, and presumably within a large 460 km-diameter crater that exposures the mantle or lower crust on Vesta.
P31A-06
Chemical Composition of Martian Rocks
In situ analyses of martian surface rocks (and soils) provided data about the chemical composition of several landing sites. One of the used techniques is the alpha-induced x-ray emission applied by the Alpha Particle X-Ray Spectrometer (APXS) onboard the current Mars Exploration Rovers (MER) Spirit and Opportunity and onboard the preceding Mars Pathfinder Rover Sojourner (MPF Mission). These measurements encompass the determination of major, minor, and (for the MER APXS) trace elements, such as Ni, Zn, and Br, as well as Cu, Pb, Sr, Y, Ga, and Ge. The obtained data indicate a remarkable compositional difference between the rocks at the different landing sites, whereas most soils including those measured by the Viking landers are chemically similar. Initially, the only chemical data of Mars were obtained by the study of a class of meteorites that turned out to be martian, which was furthermore confirmed by the discovery of a rock (by rover Opportunity) that is chemically related to those meteorites. The rocks at the Pathfinder landing site turned out to be richer in Si and K than the martian meteorites and all rocks encountered at the MER sites. At Gusev crater (the first MER landing site), two geological regions were encountered along the rover Spirit's traverse: the plains and the hills. Rocks in the plains resemble primitive basalts, while rocks located in the Columbia Hills revealed different types. Several rock classes could be cataloged based on their chemical composition. Most of the hills rocks are significantly weathered and enriched in mobile elements, such as P, Zn, S, Cl, and Br. On the other hand, a suite of ultramafic rocks was discovered for the first time on Mars. The rocks at Meridiani Planum (the second MER landing site) are salt-rich siliciclastic sediments. All rocks showed much higher S contents than the soils. High concentrations of Cl and Br were also discovered at various samples. Huge quantities of spherules were found on top of soils and outcrops along the rover's traverse. APXS measurements revealed that these spherules contain high amounts of iron that is mainly present as the mineral hematite (determined by Mössbauer spectrometry). The formation of hematite is typically, but not exclusively, an indicator for aqueous activities under oxidizing conditions. The in situ measurements at both MER landing sites point to a variety of sedimentary processes and various types of alteration processes; hence, they show clear evidence of ancient aqueous environments that discontinued long time, ago. The combination of in situ measurements and element correlations obtained by the martian meteorites implies an ancient basaltic crust with high abundances of incompatible elements (K, Rb, Nd, U, and Th) and volatile elements (S, Cl). Compared to the Earth's mantle, the martian mantle contains about twice as much Fe, is richer in moderately volatile elements like K, and has a much higher abundance of phosphorus. In conjunction with chemical data obtained from orbit, such as gamma-ray spectrometry carried out by the Mars Odyssey spacecraft, a global estimation of the composition of the martian surface is obtained and, furthermore, crustal composition can be derived.
P31A-07
TES and GRS Compositions of the Martian Surface: Evidence for Igneous and Secondary Chemical Fractionation Processes.
TES and GRS provide unique and complementary insights into martian surface compositions. TES measures the composition of the upper hundred microns of the surface while GRS measures the upper few tens of centimeters. We examine TES oxide abundances of low-albedo surfaces and compare distributions to GRS element abundances to constrain the relative roles of igneous and alteration processes on Mars. The bulk variability of compositions measured by TES is accounted for by two spectral endmembers (ST1 and ST2). TES oxide abundances (wt. %) (SiO2, Na2O, K2O, CaO, MgO, FeO, Al2O3) are calculated by combining compositions of spectral endmembers in proportion to their relative modeled abundances (vol. %). Recent GRS studies report Si, K, Fe, Th, and K/Th for 'regions' dominated by TES ST1 (RT1) and ST2 (RT2) materials. The most significant TES chemical trends are higher abundances of FeO for ST1 (ST1 15.2 % vs. ST2 12.4 %) and higher abundances of SiO2 for ST2 (ST2 57.9 % vs. ST1 53.9 %). Results from OMEGA are in agreement with TES oxides. OMEGA pyroxene maps are closely correlated with the distribution of ST1 (High FeO) while ST2 materials lack evidence of mafic bands and are consistent with an enrichment of high-silica phases. GRS RT2 chemistries have higher abundances of FeO (RT2 20.1 % vs. RT1 17.6 %), K, and Th compared to RT1. Abundances of SiO2 (RT1 44.7 % and RT2 45.8 %) and K/Th ratio do not show significant spatial variations. Chemical trends from TES and GRS appear to be in disagreement. TES ST1 is enriched in FeO while GRS RT1 is depleted in FeO. TES ST2 is enriched in SiO2 while GRS RT2 shows no enrichment in SiO2. One can account for these discrepancies, and constrain igneous and alteration processes, by considering the sampling depth differences between TES and GRS. The constant K/Th ratio across RT1 and RT2 is not consistent with subaqueous or deep subaerial aqueous weathering of basalt as K would fractionate from Th. Fractional crystallization and subduction zone magmatism could enrich K and Th, however GRS does not detect an enrichment of Si as would be expected. The lack of any significant enrichment in SiO2 between GRS RT1 and RT2 indicates that evolved volcanics (andesites) are not present in high-abundances within the upper few tens of centimeters at global scales. The favored model from the GRS team is thus initial bulk differentiation processes on Mars producing compositionally distinct magma source regions in the mantle. RT1 and RT2 basaltic provinces with distinct trace element compositions could then be produced. However, the differences in SiO2 between TES ST1 and ST2 must be taken into consideration. Thin coatings or rinds of secondary high-silica phases (tens of microns) significantly affect the shape and position of absorptions in thermal emission spectra of basalt. Such coatings on Mars may form from near-surface ice and/or surface-atmosphere interactions with little to no water penetrating or cycling into the surface. Limited degrees of alteration in only the upper few tens of microns of the surface could affect TES derived chemistries and be undetectable to GRS due to a deep sampling depth. GRS and TES chemistries support: 1) Distinct magma source regions and basaltic compositions for ST1-RT1 and ST2-RT2 and 2) Thin secondary coatings or rinds of amorphous high-silica phases on ST2-RT2 basalt.
P31A-08
The Crusts of Mars and Earth
The differentiation of terrestrial planets and large moons results in crusts with compositions differing greatly from primitive mantles. Typically, large fractions of incompatible elements, including heat-producing elements, are transferred into the crust. Mechanisms and timing of this process differ greatly from planet to planet. Accordingly, in order to understand planetary evolution, it is necessary to understand the composition and evolution of planetary crusts. Crustal evolution on Earth is perhaps the least representative of the terrestrial planets and large moons of the solar system. Although Earth substantially melted after the giant impact that resulted in the Moon, there is little evidence for the existence of a primary crust suggesting that such crust was recycled and mixed into the mantle during the Hadean. Instead, Earth has a very young, continually recycled basaltic secondary (oceanic) crust and an andesitic tertiary (continental) crust, unique in the solar system, that grew episodically over 4 Gyr, but with an average age of about 2 Gyr. The continental - oceanic crust dichotomy, temporal changes in continental crust composition, role of plume volcanism and continental growth are largely consequences of evolving plate- tectonic processes. Mars provides a valuable comparison to Earth because it is a planet that is, in many ways, intermediate between Earth and planetary bodies, such as the Moon and Mercury, that completed crustal development by about 3 Gyr and have been dormant since. Martian crust is mostly ancient (>3.5 Gyr) but volcanism has persisted, possibly episodically, to 200 Myr or younger. Proposals of early plate tectonics persist, but the weight of evidence suggests Mars is a one-plate planet. The 50 km thick crust constitutes 3.2% of the mass of the planet and, even with modest levels of LILE enrichment (K=0.33%), has had well in excess of 50% of incompatible elements removed from the mantle during early differentiation that likely resulted in a primary crust. Secondary crust in the form of young basalts (e.g., SNC meteorites) is derived from ultra-depleted mantle sources and orbital GRS data are consistent with secular variations in crustal composition.