Long-term transient simulations of past climates with comprehensive Earth System Models (ESMs) are still rare due to computational limitations and the effort of including additional processes into ESMs that are important for transient climates (e.g. processes associated with changing ice sheets, changes in sea-level, etc.). In this talk, I will present the progress within the German paleo climate modeling initiative PalMod and show results from transient simulations of the last deglaciation with the Max Planck Institute Earth System Model (MPI-ESM). The MPI-ESM simulations with prescribed ice sheets from reconstructions allow to look at changes in different compartments of the climate system and to explore differences that arise from uncertainties in different ice-sheet reconstructions. Furthermore, I will show preliminary results from first fully coupled simulations with interactive ice-sheets that are currently underway.

A recent internationally collaborative assessment (Sherwood et al. 2020) has obtained a new and narrower range of equilibrium climate sensitivity, by scrutinising key lines of evidence to see how consistent they are with high or low sensitivities. This talk will give a brief overview of that assessment and its findings, then close with perspectives on how paleoclimate studies could further narrow the uncertainties, by improving our knowledge of past forcings and interpretation of evidence from paleoclimates, but also by testing assumptions needed to interpret other lines of evidence.

Here we use one stalagmite δ18O record in Hubei province, central China to reconstruct a history of summer monsoon strength over the past 5,000 years, and focus a causal link between monsoon rainfall and human civilization. Since the publication of the Hulu records (Wang et al, 2001), an interpretation of the cave δ18O has been long and ongoing debated issue. Based on the geochemical principle and monsoon dynamics, we have suggested two mechanisms to explain the Hulu δ18O record: (i) seasonal variation of rainfall δ18O related to the position of the sub-tropical jet; and (ii) different fractions of rainout integrated between the tropical sources and the cave sites (Rayleigh fractionation). A number of modeling studies have provided theoretical confirmation for the two mechanisms. In particular, a newly-developed model well reproduces the amplitude (~5 per mil) of the Hulu δ18O record throughout the last deglacial period. On the basis of the modelling studies and the two physical mechanisms, we concluded that Chinese cave δ18O (including a precessional cycle-dominated record over the past 640ka) does represent large scale rainfall variability, but due to climate dynamics, instead of local amount effect. As our new high-resolution δ18O (0.4 years) record with extensively dating control (a total of 70 dates) that can be approximately duplicated by other caves in China as well as in Indus, we provide a new insight into the relationship of monsoon rainfall and human civilization over the past 5,000 years, a heating topic now that includes the timing and causes of Xia Dynasty (~4,000 years ago) and collapse of Tang Dynasty (~1000years ago). Indeed, we find that several civilization events agree well with timing of decadal-scale drought and/or flooding monsoon extremes, including the collapse of Neothilic culture at 4,180 years ago and the beginning of Xia Dynasty, as well as others. Careful inspections on all of the alternations of dynasties throughout the 4,000 years, we found that many, but not all, collapses of civilizations are temporally related to decadal-drought events within dating uncertainties. It is therefore induced that climate change was likely a tipping point, but did not play a key role in the birth time and fate of paleo-civil. By comparing with the ENSO signal integrated from tree-ring data and model studies, we suggested that the decadal-scale droughts commonly occurring in our record were linked to the strongly coupling of tropic oceanic/atmospheric circulations.