A coupled carbon-silicon cycle model over Earth history: Reverse weathering as a possible explanation of a warm mid-Proterozoic climate

Abstract

The balance between carbon outgassing and carbon burial controls Earth's climate on geological timescales. Carbon removal in carbonates consumes both atmospheric carbon and ocean carbonate alkalinity sourced from silicate weathering on the land or seafloor. Reverse weathering (RW) refers to clay-forming reactions that consume alkalinity but not carbon. If the cations (of alkalinity) end up in clay minerals rather than in the carbonates, carbon remains as atmospheric CO2, warming the climate. Higher silicate weathering fluxes and warmer temperatures are then required to balance the carbon cycle. It has been proposed that high dissolved silica concentrations resulting from the absence of ecologically significant biogenic silica precipitation in the Precambrian drove larger RW fluxes than today, affecting the climate. Here, we present the first fully coupled carbon-silica cycle model for post-Hadean Earth history that models climate evolution self-consistently (available as open source code). RW fluxes and biogenic silica deposition fluxes are represented using a sediment diagenesis model that can reproduce modern conditions. We show that a broad range of climate evolutions are possible but most plausible scenarios produce Proterozoic warming (+5 K relative to without RW), which could help explain the sustained warmth of the Proterozoic despite lower insolation. RW in the Archean is potentially more muted due to a lower land fraction and sedimentation rate. Key model uncertainties are the modern reverse weathering flux, the rate coefficient for RW reactions, and the solubility of authigenic clays. Consequently, within the large uncertainties, other self-consistent scenarios where Proterozoic RW was unimportant cannot be excluded. Progress requires better constraints on parameters governing RW reaction rates including explicit consideration of cation-limitations to Precambrian RW, and perhaps new inferences from Si or Li isotopes systems.