Investigating the Early Atmospheres of Earth and Mars through Rivers, Raindrops, and Lava Flows

The discovery of a habitable Earth-like planet beyond our solar-system will be remembered as one of the major breakthroughs of 21st century science, and of the same magnitude as Copernicus' heliocentric model dating from the mid 16th century. The real astrobiological breakthrough will be the added results from atmospheric remote sensing of such planets to determine habitability. Atmospheres, in both concentration and composition are suggestive of processes occurring at the planetary surface and upper crust. Unfortunately, only the modern Earth's atmosphere is known to be habitable. I investigate the density and pressure of our planet's early atmosphere before the rise of oxygen 2.5 billion years ago, because our planet was very much alive microbially. Such knowledge gives us another example of a habitable atmosphere. I also investigates the atmosphere of early Mars, as geomorphic signatures on its surface are suggestive of a past where liquid water may have present in a warmer climate, conditions suitable for the emergence of life, compared with today's 6 mbar CO2-dominated atmosphere. Using tools of fluvial geomorphology, I find that the largest river-valleys on Mars do not record a signature of a sustained hydrological cycle, in which precipitation onto a drainage basin induces many cycles of water flow, substrate incision, water ponding, and return to the atmosphere via evaporation. Rather, I conclude that while episodes of flow did occur in perhaps warmer environments, those periods were short-lived and overprinted onto a dominantly cold and dry planet. For Earth, I develop a new method of investigating atmospheric density and pressure using the size of raindrop imprints, and find that raindrop imprints preserved in the 2.7 billion year old Ventersdorp Supergroup of South Africa are consistent with precipitation falling in an atmosphere of near-surface density < 2 kg/m3 and probably > 0.1 kg/m3, compared to a modern value of 1.2 kg/m3, further suggesting a nitrogen level of at most twice present levels and perhaps well below present levels. To constrain this further, I re-evaluate a published paleobarometry technique using the vesicle size-distribution in simply emplaced lava flows and apply it to sea-level erupted lava flows from the 2.7 billion year old Fortescue group of Western Australia. Results from three flows suggest a range for atmospheric pressure 0.07 < Patm < 0.64 atm, which has profound consequences for our interpretation of the history of the nitrogen cycle by implying that the development of the nitrogenase enzyme necessary for nitrogen fixation happened very early on in the development of life.