Inferring effective radiative forcing from global model integrations

The first goal of RFMIP is to investigate the effective radiative forcing (ERF) across climate models and to understand the reasons for inter-model differences. The effective radiative forcing includes rapid adjustments in the troposphere such as the changes in clouds and vertical temperature structure, in addition to the well-known stratospheric adjustment that occurs in response to an abrupt change in greenhouse gas forcing. Rapid adjustments are distinct from climate feedbacks, which scale with changing near-surface air temperature. Rapid adjustments, including those driven by cloud changes, can be determined using radiative kernels.

As a minimum, a group of six 30-year fixed sea-surface temperature model runs will be requested from each modelling group (table 1). ERF will be diagnosed as the difference in net top-of-atmosphere radiative flux between each experiment and the model’s pre-industrial control run. Fixing sea-surface temperatures and sea-ice conditions suppresses climate feedbacks and allows ERF to be diagnosed as the difference in top-of-atmosphere (TOA) radiation fluxes between each experiment and the piClim-control run (Forster et al., 2016).

Sea surface temperatures and sea ice from a monthly climatology covering at least 30 years of a pre-industrial control run of the same model are to be used for all time slice and transient integrations. These should be prescribed according to the AMIP protocols, whereby interpolated daily data are generated preserving the prescribed monthly averaged field.

To understand the evolution of ERF over the 1850-2014 period, we also request transient simulations where forcing agents (greenhouse gases, aerosols, land-use changes, solar and volcanoes; table 2) vary to match the best-estimate historical forcing. Again, sea-surface temperatures and sea-ice are to be fixed at the pre-industrial level, to diagnose ERF as the difference of TOA fluxes. Beyond 2014, forcing timeseries are to be taken from SSP2-4.5.

In RFMIP-ERF-HistAer and RFMIP-ERF-Aer, ozone concentrations should be fixed at pre-industrial values. In concentration-based models, the concentrations of ozone should be fixed, and in emissions-based models, the ozone in the radiation code should be fixed. As solar and volcanic forcings affect stratospheric ozone, this effect will be included in the hist-nat integration of DAMIP. To get a consistent ERF estimate modelling groups should include the same dataset in the RFMIP-ERF-HistNat experiment. David Plummer, in collaboration with Michaela Hegglin, has kindly agreed to develop prescribed ozone datasets for DAMIP hist-nat, which account for solar and volcanic effects and that data will be published on input4MIPs in due course. Therefore, we suggest that modelling centres using models without coupled chemistry wait until the prescribed ozone datasets are available before commencing these experiments.

Table 1: Present-day time-slice simulations

Experiment title CMIP6 label Experiment description Years Major purposes
RFMIP-ERF-PI-Control piClim-control Pre-industrial conditions 30 Baseline for model-specific effective radiative forcing (ERF) calculations
RFMIP-ERF-Anthro piClim-anthro Present-day (2014) anthropogenic forcing (greenhouse gases, aerosols, ozone and land-use) 30 Quantify present day (2014) total anthropogenic ERF
RFMIP-ERF-GHG piClim-ghg Present-day (2014) non-ozone greenhouse gases 30 Quantify present-day (2014) ERF by greenhouse gases
RFMIP-ERF-Aer piClim-aer Present-day (2014) aerosols (ozone fixed at pre-industrial concentrations) 30 Quantify present-day (2014) ERF by aerosols
RFMIP-ERF-LU piClim-lu Present-day (2014) land-use 30 Quantify present-day (2014) ERF by land-use changes
RFMIP-ERF-4xCO2 piClim-4xCO2 CO2 concentrations set to 4 times pre-industrial 30 Quantify ERF of 4 × CO2


Table 2: Historical (1850-2014)/SSP2-4.5 (2015-2100) transient simulations

Experiment title CMIP6 label Experiment description Start End Major purposes
RFMIP-ERF-HistAll piClim-histall Time-varying forcing from all agents 1850 2100 Diagnose transient ERF from all agents
RFMIP-ERF-HistNat piClim-histnat Time-varying ERF from volcanoes, solar (including spectral) variability, etc. 1850 2100 Diagnose transient natural ERF
RFMIP-ERF-HistGHG piClim-histghg Time-varying ERF from non-ozone greenhouse gases 1850 2100 Diagnose transient ERF from greenhouse gases
RFMIP-ERF-HistAer piClim-histaer Time-varying ERF from aerosols (ozone fixed at pre-industrial concentrations) 1850 2100 Diagnose transient ERF from aerosols