Minimum simulations required: to be done as soon as you can
- Run a pre-industrial run
- Run a LGM (21ka) run with PMIP4 boundary conditions if possible (Masa et al., GMD, 2018), or as closed as PMIP4 ones otherwise (changes of orbital parameters, greenhouse gases and ice sheets). If possible, use boundary conditions consistent with the deglaciation protocol (Ruza et al., GMD, 2016) to allow a smooth transition when running the deglaciation.
For the carbon cycle CO2:
1. If possible: run simulations with freely evolving atmospheric CO2 for the carbon cycle part (NOT for the radiative part) -> this is the preferred solution
2. if 1 is not possible: run simulations with fixed atmospheric CO2 for the carbon cycle part (the CO2 value should be the same as for the radiative part)
3. even better: run both (1 and 2), especially for EMICs
- No change of code between LGM and PI, only change of boundary conditions.
- For the LGM run, also increase salinity, nutrients and alkalinity to account for the volume change between LGM and PI (unless it is already accounted for)
- No sediment nor permafrost if possible in the basic LGM run (as most models won't have them)
- For the C14: fixed atmospheric concentration
- netCDF files
- cmip6 variable names (see table below)
- Mean of the last 100 years