User Tools

Site Tools


This is an old revision of the document!

  Go to ⇒ [ PI ] - [ 6ka ] - [ 21ka ] - [ LM ]

21ka Experimental Design

Summary table

The changes listed in this table are with respect to the PI configuration.

PMIP3 Minimum solution
Orbital parameters [ ecc = 0.018994 ] - [ obl = 22.949° ] - [ peri-180° = 114.42° ]
Date of vernal equinox March 21 at Noon
Trace gases [ CO2 = 185 ppm ] - [ CH4 = 350 ppb ] - [ N2O = 200 ppb ]
[ CFC = 0 ] - [ O3 = same as in PI ]
Aerosols Same as in CMIP5 PI (see Dust forcing note below)
Solar constant Same as in CMIP5 PI
Vegetation Same as in CMIP5 PI
Ice sheets Ice sheet extent and related changes in topography provided (see below)
Topography and coastlines Land-sea mask and topography changes provided (see below) Minimum changes (see below)
Ocean bathymetry Up to groups depending on the flexibility of the model
River outflow Modified according to a river pathway map (provided, see below) Same as in CMIP5 PI
Ice sheet mass balance Add excess LGM freshwater to ocean (see below) Same as in CMIP5 PI
Mean ocean salinity +1PSU everywhere
(to be added once at the beginning of the simulation)
Mean atmospheric surface pressure Global average equal to PI value (see below)

Boundary conditions


Note that insolation should follows PMIP requirements. Please check it carefuly using the following tables (LGM insolation tables)

Check carefully the date of the Vernal equinox, because it has implications to compare the paleo and PI seasonal cycles

The ice sheet provided for PMIP3/CMIP5 LGM experiments is a blended product obtained by averaging three different ice sheets: ICE-6G v2.0 provided by Dick Peltier, MOCA provided by Lev Tarasov and ANU provided by Kurt Lambeck. More information about this is available in the LGM icesheet description page.

Implementing the LGM ice-sheets implies changing also the land-sea mask and the topography as described below.

Land-sea mask

This mask should be prescribed using the data provided in where the sftlf variable is 0 (%) over the ocean and 100 (%) over land.
If this proves difficult for a modelling group, the minimum implementation is:

  • to fill in the Hudson Bay and the Barents sea, which were covered by the main northern hemisphere ice-sheets.
  • to close the Bering Strait (by changing the land-sea mask or imposing zero flux on water and sea-ice across the Bering Strait).
  • to make sure that the ocean model configuration keeps the following straits open:
    • Denmark Strait (between Greenland and Ireland).
    • Gibraltar Strait.
    • Indonesia Strait (Makassar Strait?).


The change in surface elevation specified by the orog_diff variable provided in should be added to the present-day surface elevation, so that all the participating models will be perturbed in the same way.
Note: this specified change assumes that the surface elevation over present-day oceans is zero (m).

Ice-sheet extent (aka 'land-sea mask')

The ice-sheet extent should be prescribed using the data provided in where the sftgif variable is 0 (%) where there is no ice-sheet or 100 (%).
This mask includes ice shelves, which should be included in models that allow this feature.

  • The ice shelves can be defined from the sftgif and sftlf variables as areas where:
    sftgif == 100 and sftlf == 0
  • If you implement the provided ice shelves, you should keep the Greenland-Iceland ridge open in the ocean configuration.

Data files are available on the Files for PMIP3 experiments page.

Land-sea mask

The land-sea mask is provided by the ICE6G model, except for Antarctica where you should deduce it from the averaged ice-sheet.

Data files are available on the Files for PMIP3 experiments page.

Note: the minimum change is to close the Bering Strait by changing the land-sea mask or imposing zero flux on water and sea-ice across the Bering Strait.

River outflow

The river pathways and basins should be at least adjusted so that fresh water is conserved at the Earth's surface: care should be taken that rivers reach the ocean (due to the lower sea level at LGM, some river mouths have to be displaced towards the coast).

You can use the river routing provided by Lev Tarasov to change the routing scheme in your code, so that the river pathways is consistent with the presence of the ice-sheet.

Data files are available on the Files for PMIP3 experiments page.

Ice-sheet mass balance

It is advised to ensure a closed fresh water balance at the Earth's surface: snow accumulating on the ice-sheets should be redistributed to the oceans, either globally or in the adjacent oceans (see PMIP2 recommendation mass balance recommendation).


The vegetation should be treated as in the CMIP5 PI experiment. The reason is that in CMIP5 we test the version of the model used for future climate projections. Since OA and ESM models will be considered, depending on the model used the vegetation will be

  • prescribed to PI (which means both vegetation types and LAI are prescribed )
  • Prescribed to PI with interactive LAI (models with interactive carbon cycle, but no vegetation dynamics)
  • Computed by the model (models with dynamic vegetation)

For Earth System Models with interactive carbon cycle

The simulations should be forced by the prescribed LGM CO2 concentrations. Please use the same protocol as in CMIP5 to store the diagnostic carbon fluxes and the variables needed for PCMIP (see pcmip Project and CMIP5)

Dust forcing

Some ESM have interative aerosols. In that case compute dust and associated forcing online, as in PI. If this is not the case then the recommendation is to keep dust and aerosols as in PI (I.E no change for 21 ka).

Initial conditions

The surface pressure field must be adjusted to the change in surface elevation over the continents. This can be done:

  • either by gradually changing the surface elevation in order to avoid generating gravity waves,
  • or by adjusting the initial pressure field to the LGM surface elevation.

If you choose the second option, you must be careful to conserve atmospheric mass.

The spin up procedure is up to group, following CMIP5 approach or from a previous cold state

Note several groups share the same ocean models for which an initial state can be provided from PMIP2 experiments.

If you need an initial state from a other group, you can contact

Groups with high resolution model for which it is too difficult to run long simulations should contact to find the best alternative solution.

[ PMIP3 Wiki Home ] - [ Help! ] - [ Wiki syntax ]   - [ Top ]

pmip3/design/21k/final.1269600937.txt.gz · Last modified: 2010/03/26 10:55 by jypeter