Transient climate simulations of the deglaciation 21-9 thousand years before present (version 1) - PMIP4 Core experiment design and boundary conditions

• Published in: Geoscientific Model Development
• Main Authors: Ivanovic, Ruza, Gregoire, Lauren, Kageyama, Masa, Roche, Didier, Valdes, Paul, Burke, Andrea, Drummond, Rosemarie, Peltier, W. Richard, Tarasov, Lev
• Format: Article in Journal/Newspaper
• Language: English
• Published: 2016
• Subjects: Ice Sheet
• Online Access: https://research-portal.st-andrews.ac.uk/en/publications/a45be17e-3cb5-48da-a2bc-f948a9b0a24f; https://doi.org/10.5194/gmd-9-2563-2016; https://research-repository.st-andrews.ac.uk/bitstream/10023/9300/1/Burke_2016_GMD_Deglaciation_CC.pdf

The last deglaciation, which marked the transition between the last glacial and present 23 interglacial periods, was punctuated by a series of rapid (centennial and decadal) climate 24 changes. Numerical climate models are useful for investigating mechanisms that underpin the climate change events, especially now that some of the complex models can be run for multiple millennia. We have set up a Paleoclimate Modelling Intercomparison Project (PMIP) working group to coordinate efforts to run transient simulations of the last deglaciation, and to facilitate the dissemination of expertise between modellers and those engaged with reconstructing the climate of the last 21 thousand years. Here, we present the design of a coordinated Core experiment over the period 21-9 thousand years before present (ka) with time varying orbital forcing, greenhouse gases, ice sheets, and other geographical changes. A choice of two ice sheet reconstructions is given, and we make recommendations for prescribing ice meltwater (or not) in the Core experiment. Additional focussed simulations will also be coordinated on an ad-hoc basis by the working group, for example to investigate more thoroughly the effect of ice meltwater on climate system evolution, and to examine the uncertainty in other forcings. Some of these focussed simulations will target shorter durations around specific events in order to understand them in more detail and allow the more computationally expensive models to take part.