This is an old revision of the document!
PMIP3 WG - Deep-time Model-data Intercomparison Project - "DeepMIP"
Questions to be addressed
What are the key scientific questions?
GF: Efficacy of fast feedbacks in a warm world? What feedbacks that operate in warm climates that done operate in more equitable climes? E.g. are their hidden feedbacks and/or sudden jumps in feedback efficacy that lie in wait as we warm the world toward Eocene conditions? Background climate state dependency of feedbacks etc.
YD: It looks like we are unable to melt the Antarctic ice-sheet with reasonable CO2 levels during the Pliocene (unless tuning all the parameters to get it). Conversely, CO2 is sometimes very low before the E-O transition, what does it mean ? Why do we not see changes in d180, I am not convinced that the small geographical changes characterizing the early Cenozoic could explain a lower Earth sensitivity to glaciation.
DN: What is the impact of a warmer climate on biogeochemical cycles (and weathering)? How did the hydrological cycle look like in a warmer world?
What time periods should we target?
GF: We need a big signal and lots of data coverage. Time periods that are easy to spot in the record and have lots (ish) of data already are Eocene climatic optimum, middle Eocene climatic optimum, PETM.
GL: As a focus, I suggest two time intervals: Eocene and Miocene.
BB: Oligocene and Miocene evolution of climate, ice sheets and the carbon cycle.
DN: Miocene (maybe MCO), E/O boundary, ECO, PETM.
What is needed for experimental design – paleogeographies and CO2 primarily – what datasets can we use?
MH: a full set of early Eocene boundary conditions submitted to Geoscientific Model Development which will be freely available (Nick Herold is spearheading this). It will include topo,bathymetry,veg, tidal dissipation, and aerosol forcing files at 1°x1° among other parameters. the DEM is based on a merging of Markwick's topo and Mueller's bathy/rotations…. it should include everything necessary setting up a coupled GCM run (nothing is turnkey of course, but all the fundamentals are there). hopefully it might provide a basis for future intercomparison efforts and provide a baseline for subsequent modification or comparison. the necessary scripts for rotation the topo/veg to earlier and later time intevals (say 65 and 35mya) will also be provided… Once the article is submitted, we will make it available for perusal and criticism…
What are the best datasets for model-data comparisons, and how can these be improved?
What are the best methodologies for model-data comparisons (close collaboration with the Past2Future working group)
GF: I think a radical way to do this would be to move away from simple difference to modern type studies that have been done to date, instead we could try to look across a number of natural warming events and see whether the models capture the warming the geological record shows. So for instance, pre, peak and post PETM, MECO and EECO - do the models show similar trajectories to the data? CO2 would be much better constrained this way and so will the SSTs (e.g. most of the TEX86 uncertainty in your Eocene paper is due to calibration, for these shorter time periods you could use delta T change which is largely independent of calibration).
How can we fund this?
Bas de Boer
Tom Dunkley Jones
Anna von der Heydt
Sze Ling Ho
Sandy Kirtland Turner
David De Vleeschouwer
(very) informal meeting at AGU 2013.
(very) informal meeting at Namur, Belgium, 2014.
NCAR, Boulder, Colorado. 14th-15th January 2016.
Freely available Paleogeographies