The Snowball Stratosphere

Abstract

Key Points:• The simulated snowball stratosphere with or without ozone displays weaker circulation than the modern case. • The snowball stratosphere with ozone has weaker wave forcing and a stronger jet than the modern and has no sudden stratospheric warmings. • Changes to stratospheric circulation during the snowball do not impact pCO 2 estimated from proxy data unless pCO 2 was much greater than pO 2 .-1-arXiv:1909.12717v1 [physics.ao-ph] AbstractAccording to the Snowball Earth hypothesis, Earth has experienced periods of low-latitude glaciation in its deep past. Prior studies have used general circulation models (GCMs) to examine the effects such an extreme climate state might have on the structure and dynamics of Earth's troposphere, but the behavior of the stratosphere has not been studied in detail.Understanding the snowball stratosphere is important for developing an accurate account of the Earth's radiative and chemical properties during these episodes. Here we conduct the first analysis of the stratospheric circulation of the Snowball Earth using ECHAM6 general circulation model simulations. In order to understand the factors contributing to the stratospheric circulation, we extend the Statistical Transformed Eulerian Mean framework. We find that the stratosphere during a snowball with prescribed modern ozone levels exhibits a weaker meridional overturning circulation, reduced wave activity, stronger zonal jets, and is extremely cold relative to modern conditions. Notably, the snowball stratosphere displays no sudden stratospheric warmings. Without ozone, the stratosphere displays slightly weaker circulation, a complete lack of polar vortex, and even colder temperatures. We also explicitly quantify for the first time the cross-tropopause mass exchange rate and stratospheric mixing efficiency during the snowball and show that our values do not change the constraints on CO 2 inferred from geochemical proxies during the Marinoan glaciation (∼635 Ma), unless the O 2 concentration during the snowball was orders of magnitude less than the CO 2 concentration.

Comparison between ozone column depths and methane lifetimes computed by one- and three-dimensional models at different atmospheric O 2 levels

“…Because the stratospheric polar vortex forms due to latitudinal temperature gradients, when O 3 is reduced, the average velocity of the polar vortex decreases. This change in wind fields seen in the WACCM6 simulations, including a weakened Brewer–Dobson circulation, is consistent with previous work [44,45] and affects the latitudinal distribution of O 3 (more details are included in the electronic supplementary material). The other 3-D models also have sophisticated vertical transport schemes, but the fact that they do not all agree shows that this problem is complicated.Other factors that contribute to differences in calculated ozone column depths include: the temperature structure which modulates reactions rates and densities; chemical mechanism differences (e.g.…”

Hydrological Cycle Changes Explain Weak Snowball Earth Storm Track Despite Increased Surface Baroclinicity

“…Snowball Earth by imposing ice everywhere (ocean covered with sea ice and land covered with glaciers) in the modern simulation with an ice albedo of 0.6. The AGCM simulations are identical to those in Graham et al (2019). Climate change in the AGCM is the difference of the Snowball Earth and modern simulations.…”

“…All of the mechanisms are based on dry zonally asymmetric dynamics; for example, they appeal to changes in stationary wave amplitude (Hall et al, 1996), reflection (Lofverstrom et al, 2016), and structure (Riviere et al, 2018) due to the orographic forcing of the continental ice sheets or changes in upstream seeding (Donohoe & Battisti, 2009). Zonal asymmetries are weaker in Snowball Earth simulations (Graham et al, 2019); however, the weakening of the NH wintertime zonal mean Snowball Earth storm track is too large (∼4 PW, see Figure 7 in Pierrehumbert, 2005) to be explained by zonally asymmetric LGM mechanisms. Thus, alternative mechanisms are needed.…”