Abstract.One of the critical issues of the Snowball Earth hypothesis is the CO 2 threshold for triggering the deglaciation. Using Community Atmospheric Model version 3.0 (CAM3), we study the problem for the CO 2 threshold. Our simulations show large differences from previous results (e.g. Pierrehumbert, 2004Pierrehumbert, , 2005Le Hir et al., 2007). At 0.2 bars of CO 2 , the January maximum near-surface temperature is about 268 K, about 13 K higher than that in Pierrehumbert (2004Pierrehumbert ( , 2005, but lower than the value of 270 K for 0.1 bar of CO 2 in Le Hir et al. (2007). It is found that the difference of simulation results is mainly due to model sensitivity of greenhouse effect and longwave cloud forcing to increasing CO 2 . At 0.2 bars of CO 2 , CAM3 yields 117 Wm −2 of clear-sky greenhouse effect and 32 Wm −2 of longwave cloud forcing, versus only about 77 Wm −2 and 10.5 Wm −2 in Pierrehumbert (2004, 2005), respectively. CAM3 has comparable clear-sky greenhouse effect to that in Le Hir et al. (2007), but lower longwave cloud forcing. CAM3 also produces much stronger Hadley cells than that in Pierrehumbert (2005).Effects of pressure broadening and collision-induced absorption are also studied using a radiative-convective model and CAM3. Both effects substantially increase surface temperature and thus lower the CO 2 threshold. The radiativeconvective model yields a CO 2 threshold of about 0.21 bars with surface albedo of 0.663. Without considering the effects of pressure broadening and collision-induced absorption, CAM3 yields an approximate CO 2 threshold of about 1.0 bar for surface albedo of about 0.6. However, the threshold is lowered to 0.38 bars as both effects are considered.

A compilation of 53 14 C ages on moraines constrains the Holocene glacial history of Tibet and the surrounding mountains. Several Holocene glacial stages are identified. The Little Ice Age had three substages at 1.0-0.13 cal. ka BP. However, the glacial advances may have occurred 200-600 a earlier in southern and eastern bordering mountains than in the northern bordering mountains. The earlier glacial advance might have been driven by humid conditions in bordering areas of Tibet. Most glacial advances in the Neoglaciation occurred at 3.5-1.4 cal. ka BP. They can be identified in almost every mountain range. An early Holocene glacial advance occurred 9.4-8.8 cal. ka BP, in both central Tibet and bordering mountains. The timings of Holocene glacial advances are synchronous with the cooling periods in the d 18 O record from ice cores and with the results of optically stimulated luminescence dating and cosmogenic exposure dating in Tibet.

Abstract. We identify the "hard snowball" bifurcation point at which total sea-ice cover of the oceans is expected by employing the comprehensive coupled climate model CCSM3 (Community Climate System Model version 3) for two realistic Neoproterozoic continental configurations, namely a low-latitude configuration appropriate for the 720 Ma Sturtian glaciation and a higher southern latitude configuration reconstructed for 570 Ma but which has often been employed in the past to study the later 635 Ma Marinoan glaciation. Contrary to previous suggestions, we find that for the same total solar insolation (TSI) and atmospheric CO 2 concentration (pCO 2 ), the 570 Ma continental configuration is characterized by colder climate than the 720 Ma continental configuration and enters the hard snowball state more easily on account of the following three factors: the higher effective albedo of the snow-covered land compared to that of sea ice, the more negative net cloud forcing near the ice front in the Northern Hemisphere (NH), and, more importantly, the more efficient sea-ice transport towards the Equator in the NH due to the absence of blockage by continents. Beside the paleogeography, we also find the optical depth of aerosol to have a significant influence on this important bifurcation point. When the high value (recommended by CCSM3 but demonstrated to be a significant overestimate) is employed, the critical values of pCO 2 , beyond which a hard snowball will be realized, are between 80 and 90 ppmv (sea-ice fraction 55 %) and between 140 and 150 ppmv (sea-ice fraction 50 %) for the Sturtian and Marinoan continental configurations, respectively. However, if a lower value is employed that enables the model to approximately reproduce the present-day climate, then the critical values of pCO 2 become 50-60 ppmv (sea-ice fraction 57 %) and 100-110 ppmv (sea-ice fraction 48 %) for the two continental configurations, respectively. All of these values are higher than previously obtained for the present-day geography (17-35 ppmv) using the same model, primarily due to the absence of vegetation, which increases the surface albedo, but are much lower than that obtained previously for the Marinoan continental configuration using the ECHAM5/MPI-OM model in its standard configuration (∼ 500 ppmv). However, when the sea-ice albedo in that model was reduced from 0.75 to a more appropriate value of 0.45, the critical pCO 2 becomes ∼ 204 ppmv, closer to the values obtained here. Our results are similar to those obtained with the present-day geography (70-100 ppmv) when the most recent version of the NCAR model, CCSM4, was employed.

No abstract