Billions of years ago, liquid water flowed across the surface of Mars, as indicated by the pervasive fluvial features that cover the planet's surface to this day. To stably maintain this phase of H2O, the martian atmosphere would have needed to be much, much thicker that it is today. As the atmosphere is ~95% carbon dioxide, there are billions of tons of carbon that are unaccounted for. Atmospheric physics and chemistry suggest that sequestration of CO2 into carbonate rocks would be the most likely scenario, yet a corresponding quantity of carbonate-bearing rocks has not been identified by satellite-based analyses. Except for a chunk of such rock found at Nili Fossae. In a new analysis of that outcrop and estimates of carbon quantities, a new study from Christopher Edwards and Bethany Ehlmann at Caltech lays out three possibilities: 1) the atmosphere may have been thick in the past, but it had a lower proportion of CO2 than it does today; 2) much of the carbon-bearing atmosphere was lost to space; or 3) large carbonate rock repositories - about 34 Nili Fossae equivalents - remain buried, yet to be discovered...

Full paper here, abstract below:

On Earth, carbon sequestration in geologic units plays an important role in the carbon cycle, scrubbing CO2 from the atmosphere for long-term storage. While carbonate is identified in low abundances within the dust and soils of Mars, at <1 wt% in select meteorites, and in limited outcrops, no massive carbonate rock reservoir on Mars has been identified to date. Here, we investigate the largest exposed carbonate-bearing rock unit, the Nili Fossae plains, combining spectral, thermophysical, and morphological analyses to evaluate the timing and carbon sequestration potential of rocks on Mars. We find that the olivine-enriched (∼20%–25%) basalts have been altered, by low-temperature in situ carbonation processes, to at most ∼20% Fe-Mg carbonate, thus limiting carbon sequestration in the Nili Fossae region to ∼0.25–12 mbar of CO2 during the late Noachian–early Hesperian, before or concurrent with valley network formation. While this is large compared to modern-day CO2 reservoirs, the lack of additional, comparably sized post–late Noachian carbonate-bearing deposits on Mars indicates ineffective carbon sequestration in rock units over the past ∼3.7 b.y. This implies a thin atmosphere (≲500 mbar) during valley network formation, extensive post-Noachian atmospheric loss to space, or diffuse, deep sequestration by a yet-to-be understood process. In stark contrast to Earth’s biologically mediated crust:atmosphere carbon reservoir ratio of ∼104–105, Mars’ ratio is a mere ∼10–103, even if buried pre-Noachian crust holds multiple bars.