The morphology of Mars indicates that abundant liquid water existed in its early history. Where did this water come from and where did it go? If the water was degassed from the martian mantle, we would expect to find evidence of water in igneous martian minerals. If such igneous phases show little evidence for water, that might suggest water-rich materials came as a late-stage accretion to the martian surface. If the martian crust has acted through time as a significant sink for surface water, then we might expect to find abundant hydrated minerals formed by weathering processes at the martian surface. What do martian meteorites tell us about these questions?

Martian meteorites are known to contain martian water (Figure 3), some in igneous minerals (e.g., amphibole and mica) that crystallized from water-bearing magmas, and some in weathering products. However, water identified in igneous minerals appears to exist in relatively low abundance (~0.1-0.5%), and this implies their parent magmas were low in water. It is not apparent that such concentrations are sufficient to produce the amount of degassed water required to generate martian fluvial features. Further, an analysis of the deuterium/hydrogen (D/H) isotopic ratio of water in meteorite apatite and other minerals shows large fractionated values similar to atmospheric hydrogen (see next topic), suggesting that even the water in this igneous phase may have been derived from the martian surface.

Those weathering products in martian meteorites that contain water include silicate clays, hydoxides, and various salts, all clearly formed by interaction at low temperature of liquid martian water with igneous minerals that comprise the meteorites. Several martian meteorites contain traces of precipitated minerals such as Ca-carbonate and Ca-sulfate (Figure 3). The nakhlite meteorites contain the most diverse suite of hydrated phases, including carbonates, sulfates, and layer-structured silicate clays (iron-rich smectites), often in association with ferric iron oxides and hydroxides. These mineral assemblages, as well as oxygen isotopic analyses of water extracted from the nakhlites (Figure 4), suggest that the altering water was cool by geologic standards (<100°C), possibly cold (~0°C), and strongly oxidizing.

Unlike the younger martian meteorites, 4.5 billion years old ALH84001 contains abundant inclusions of distinctively layered sequences of Ca-Mg-Fe-carbonates (Figure 5). One interpretation is that they formed in a warmer, more reducing aqueous environment than the Nakhlites-perhaps hydrothermal. Another interpretation is that they formed in a cool reducing aqueous environment. Preliminary analyses indicate that the carbonates may have been depositied in ALH84001 as early as 3.6 billion years ago. Thus, martian meteorites suggest aqueous activity may have persisted throughout martian history. Because martian meteorites never spent substantial time at the very surface, more highly weathered surface materials are likely to contain greater mineral diversity and higher concentrations of water-bearing phases, which is consistent with Viking chemical analyses of martian soils.