Modelling and reconstructing past climate change is the key to understanding Earth system dynamics at time scales beyond decades and centuries on which slow processes such as ice sheet dynamics and migration of vegetation unfold. But paleo climate science also challenges our understanding of short-term Earth system dynamics evolving under different forcing and boundary conditions. The notorious underestimation of the amplitude of tropical monsoon changes in response to different orbital forcing is one example. Syntheses of high-resolution paleo records also allow to assess the diversity of interannual to multidecadal events and their role in extreme regional conditions. Particularly interesting are tipping points, i.e., accelerated changes into a new climate and ecosystem states driven by positive feedbacks. Hence this session invites contributions from the broad spectrum of Earth system modelling of past climates and past climate change. Contributions combining model results, paleo records and explicitly modeling of past climate sensors (isotopes, ocean tracers, tree-ring, speleothems…) to test ideas related to the WCRP Grand Science Challenges are likely to have particular resonance.