The control integration, in which the atmospheric forcing is kept constant, simulates a period of 1000 years. The value of CO2 in the whole run is approximately equal to the value observed in 1958 (this is the only thing which "fixes" the model years to real years). The initial conditions in the model atmosphere and ocean are in no way related to the conditions in 1958. Therefore due to natural variability both in the model and in the observed climate the state is different even though the radiative forcing was about equal between model and nature. These data represent monthly averaged surface values of selected variables for the IPCC-Data Distribution Centre. (see also https://www.ipcc-data.org/ )