21st Century Scenario Forcing Increases More for CMIP6 Than CMIP5 Models

Although the Coupled Model Intercomparison Project 6 (CMIP6) protocol provides an experiment to estimate effective radiative forcing (ERF), it is only quantified for few models. We present new estimates of ERF for models participating in CMIP6 by applying the method developed in Fredriksen et al. (2021, https://doi.org/10.1029/2020JD034145), and validate our approach with available fixed‐SST forcing estimates. We estimate ERF for experiments with abrupt changes of CO2, 1% increase of CO2, historical forcings, and future scenarios, and demonstrate that CMIP6 ERF is lower than CMIP5 ERF at the end of the historical period, but grows faster than CMIP5 in the future scenarios, ending up at higher levels than CMIP5 at the end of the 21st century. The simulated radiative efficiency of CO2 has not changed much, suggesting that the larger future increase in CO2 concentrations in CMIP6 compared to CMIP5 is important for explaining the forcing difference. Plain Language Summary To understand climate model responses, it is useful to separate between the drivers of climate change and their responses. We present new estimates of the drivers, called the effective radiative forcing, for the latest generation of climate models (Coupled Model Intercomparison Project 6 (CMIP6)). This estimates the energy imbalance at the top of the atmosphere and is a measure of human and natural influences on climate. Normally this requires additional climate model experiments to make these estimates, but since these have only been run for a few models, we are here aiming to make the best alternative estimates based on more widely available data, following the method in Fredriksen et al. (2021, https://doi.org/10.1029/2020JD034145). We show that our forcing estimates are growing faster during the 21st century for the new CMIP6 models than for the previous generation of models (CMIP5), and suggest this can be attributed to the higher CO2 concentrations in future scenarios for CMIP6 compared to CMIP5. Key Points Our method to estimate effective radiative forcing based on common model diagnostics is consistent with Coupled Model Intercomparison Project 6 (CMIP6) fixed‐SST estimates We present coherent estimates for abrupt CO2, 1% CO2, historical, and future scenario experiments Forcing estimates for the 21st century grow faster for CMIP6 than for CMIP5 models