What caused Earth's temperature variations during the last 800,000 years? Data-based evidence on radiative forcing and constraints on climate sensitivity

TLDR: In this paper, a global cooling of 3.9 − 0.8 − 1.6 − 8.0 − 8 − 1 − 8−1.6 K was estimated for the last glacial maximum (LGM) using only the Planck feedback that parameterises the direct effect on the radiative balance.

About: This article is published in Quaternary Science Reviews.The article was published on 2010-01-01 and is currently open access. It has received 159 citations till now. The article focuses on the topics: Climate sensitivity & Climate state.

Figures

Figure 6: Dust-albedo feedback. (A) Annual and global mean local insolation I (Berger, 1978). (B) Variation in dust flux measured in EPICA Dome C (Lambert et al., 2008) on the EDC3 age scale (Parrenin et al., 2007) (left y-axis) and calculated changes in atmospheric albedoαA (right y-axis). Original data are averaged with a 1000 year running mean. (C) Perturbation in the radiative budget caused by the dust-albedo feedback. The right y-axis shows the temperature anomalies ∆TE,P of the Planck feedback.

Figure 8: Sum of radiative forcing∆Rsubtotal (left y-axis) of individual processes and its respective temp rature anomaly∆TE,P,subtotal (right y-axis) considering only the Planck feedback but without those of water vapour, lapse rate and clouds (∆Rsubtotal = ∆Rorbit + ∆RGHG + ∆Rland cryo + ∆Rsea ice + ∆Rdust + ∆Rveg). Grey and yellow shadings in∆Rsubtotal represent the lower and upper estimated uncertainty, respectively. Reconstructed temperature changes for comparison: (i) Antarctic∆TEPICA from the EPICA Dome C ice core (Jouzel et al., 2007) on the EDC3 age scale (Parrenin et al., 2007) scaled with a constant polar amplification of two to a suggested change in global surface air temperature. (ii) Deep ocean∆Tocean deconvoluted from the benthicδ18O stack LR04 (Bintanja et al., 2005; Lisiecki and Raymo, 2005).

Figure 7: A compilation of individual radiative forcings dueto orbital variation (∆Rorbit), the GHG CO2, CH4 and N2O (∆RGHG), land cryosphere including ice sheets and associated sea level change and snow cover (∆Rland cryo), sea ice (∆Rsea ice), atmospheric dust loadings (∆Rdust), and vegetation (∆Rveg). Feedbacks from water vapour, lapse rate, and clouds are omitted here.

Figure 3: Albedo feedback from the land cryosphere. (A) Annual mean local insolation at the top of the atmosphereI at 40 − 80◦ N (Berger, 1978). (B) Area changes of (i) land ice sheets (North America and Eurasia only) a d (ii) land without ice (both in40 − 80◦ N deconvoluted from the benthicδ18O stack LR04 (Bintanja et al., 2005; Lisiecki and Raymo, 2005)), and (iii)global shelf area effected by sea level change based on own assumptions. (C) Changes in surface albedo αS on areas plotted in (B), those over land ice and ice-free landare again simulation results from (Bintanja et al., 2005), while albedo change on exposed shelves are our own assumption. (D) Perturbation in the radiative budgetcaused by the albedo feedback from land ice sheets, snow cover on non-glaciated land (40− 80◦ N) and sea level change. The right y-axis shows the temperature anomalies∆TE,P of the Planck feedback. The legend is valid for sub-figures B,C, and D. The albedo over northern hemispheric land ice sheets is undefined around130 kyr, because these ice sheets disappeared completely. Ice sheets did not vanish completely in the other interglacials, thus albedo over land ice was defined during all other times.

Figure 9: (A) Probability distribution of global cooling atthe LGM. (B) Probability distribution of the LGM global radiat ve forcing relative to today from greenhouse gases, orbital forcing, ice sheets, vegetationnd dust. (C) Uncertainty distribution in the scaling factor used to translate LGM climate sensitivity to present day sensitivity. (D) Probability distribution forpresent day equilibrium climate sensitivity for atmosphericCO2 doubling resulting from panels A-C. The estimated climate sensitivity includes the Planck, water vapour, lapse rate, sea ice/snow cover albedo and cloud feedbacks oc urring on timescales of decades or less (equivalent to the climate model or Charney sensitivity) and is the quantity that is relevant for future climate projections timescales of decades to centuries. The effects of ice sheets and vegetation changes are treated as aforcing. Red solid lines show the standard case as describedin the text, while black and blue dashed lines show the distributions if all input uncertainties are increas d by 50% or reduced by 33%, respectively. For the radiative forcing in (B) the black and blue dashed lines represent the upper and lower estimated uncertainty (±1.9 W m−2 and±0.9 W m−2, respectively), and the red solid line their mean value (±1.4 W m−2).

Figure 4: Sea ice-albedo feedback. (A) Annual mean local insoation at the top of the atmosphereI at50−70◦ S and60−90◦ N, the regions of SH and NH sea ice, respectively (Berger, 1978). (B) Area changes of sea ice. Estimates are based on (i) Antarctic temperature changes for theSH (Jouzel et al., 2007) and (ii) northern hemispheric temperature changes for the NH (Bintanja et al., 2005; Lisiecki and Raymo, 2005), linearly related to LGM to present sea ice area reconstructions. (C) Assumed changes in surface albedoαS from open ocean to sea ice. (D) Perturbation in the radiativebudget caused by the sea ice-albedo feedback. The right y-axis shows the temperature anomalies∆TE,P of the Planck feedback. The legend is valid for sub-figures A,B, and D.