This is an old revision of the document!
The Preindustrial control should be the same as the one used for CMIP5 experiments. It is thus the same reference as the one for 20th century simulations and future climate projections (see taylor_cmip5_dec31.pdf).
Depending on the complexity of the model used the vegetation will be either:
Land use should be included in the PI vegetation map. In that case, care should be taken that the land use map used in the PI simulation is consistent with the one used for land use evolution either for the 20th century or the last Millennium.
The simulations are forced by the CO2 concentrations. Please use the same protocol as in CMIP5 to store the carbon fluxes and the variables needed for PCMIP (see list here External Link)
Note that insolation should follow PMIP requirements. Please check it carefully using the following tables (PI insolation tables)