How cloud radiative forcing could allow Snowball Earth deglaciation

Abstract:

Neoproterozoic, and possibly Paleoproterozoic, glaciations represent the most extreme climate events in post-Hadean Earth, and may link closely with the evolution of the atmosphere and life. According to the Snowball Earth hypothesis, the entire ocean was covered with ice during these events for a few million years, during which time volcanic CO2 increased enough to cause deglaciation. Geochemical proxy data and geochemical model calculations suggest that the maximum CO2 was 0.01–0.1 by volume, but early climate modeling suggested that the Snowball was not even close to deglaciation at CO2=0.2. Unless resolved, this discrepancy would be problematic for the Snowball Earth hypothesis.

First, I will present results from six GCMs suggesting that positive cloud radiative forcing would likely have warmed a Snowball Earth enough to reduce the CO2 required for deglaciation by a factor of 10–100. Next, I will present results from a cloud resolving model run on a small domain that are consistent with the GCM results and allow us to understand the GCM behavior better. The cloud resolving model produces convection that extends vertically to a similar temperature as modern tropical convection. This convection produces clouds that resemble stratocumulus clouds under an inversion on modern Earth, which slowly dissipate by sedimentation of cloud ice. There is enough cloud ice for the clouds to be optically thick in the longwave, and the resulting cloud radiative forcing is similar to that produced in GCMs run in Snowball conditions. This result is robust to large changes in the cloud microphysics scheme because the cloud longwave forcing, which dominates the total forcing, is relatively insensitive to cloud amount and particle size. Taken together, these results from a hierarchy of models suggest that positive cloud radiative forcing would warm a Snowball Earth enough to allow deglaciation at a CO2 consistent with geochemical data.