CFMIP has been contributing to the WCRP Coupled Model Intercoparison Project (CMIP) from its third phase (CMIP3), and is one of the important satellite MIPs in the distributed structure of CMIP6. The third phase of CFMIP (CFMIP3) experiments has two major objectives:

Objective 1: Inform improved assessments of climate change cloud feedbacks by (a) improving our understanding of cloud-climate feedback mechanisms, and (b) improving evaluation of clouds and cloud feedbacks in climate models.

Objective 2: Use the CFMIP experimental hierarchy and process diagnostics to better understand other aspects of the climate response, such as changes in circulation, regional-scale precipitation and non-linear change.

CFMIP3 family

There are three categories of GCM experiments relevant to CFMIP3 (see also summary diagram below).

  • DECK+CMIP6 historical (piControl, historical, AMIP, 1%CO2, abrupt 4xCO2): Mandatory to CMIP6.

  • Tier 1 : Required in CFMIP3

  • Tier 2 : Optional

So far, 15 modeling groups have proposed contribution to CFMIP3 using their GCMs: BCC-CSM2-MR, CanESM5, CESM2, CIESM, CNRM-CM6, GFDL-CM4, GISS-E2.1, HadGEM3-GC31-LL, IPSL-CM6, MIROC6, MPI-ESM1.2-LR, MRI-ESM2, NICAM, NorESM2, TaiESM1. They will participate via CMIP6 activity following its timescale (2017-2020), and the CFMIP3 Tier 1 experiments are expected to be published by the end of 2018, Tier 2 by mid 2019. Outputs of the CFMIP3 experiments will be available on the Earth System Grid (ESG) as a part of the CMIP6 archives.

Description of CFMIP3 experiments

Each of the CFMIP3 experiments has scientific questions to be addressed with, and can be done using either AGCM or CGCM. Here is the list of the CFMIP3 experiments.

Tier 1

AMIP and Aquaplanet runs

  • Experiment names: amip-p4K, amip-4xCO2, amip-future4K, aqua-control, aqua-p4K, aqua-4xCO2

  • Science question: What are the physical mechanisms underlying the range of cloud feedbacks and cloud adjustments predicted by climate models, and which models have the most credible cloud feedbacks?

  • Model: AGCM

  • Length of integration: 174 yrs (36 yrs each for amip and 10 yrs each for aqua)

Tier 2

AMIP minus uniform 4K SST

  • Experiment names: amip-m4K

  • Science question: Are cloud feedbacks consistent for climate warming and cooling, and if not, why?

  • Model: AGCM

  • Length of integration: 36 yrs

  • Lead PIs: Mark Webb and Bjorn Stevens

AMIP/Aquaplanet experiments with clouds transparent to longwave radiation (COOKIE LW off)

  • Experiment names: amip-lwoff, amip-p4K-lwoff, aqua-control-lwoff, aqua-p4K-lwoff

  • Science question: How do cloud-radiative effects impact the structure, the strength and the variability of the general atmospheric circulation in present and future climates?

  • Model: AGCM

  • Length of integration: 92 yrs (36 yrs each for amip and 10 yrs each for aqua)

  • Lead PIs: Sandrine Bony and Bjorn Stevens

Abrupt +/-4% solar forcing

  • Experiment names: abrupt-solp4p, abrupt-solm4p

  • Science question: How do responses in the climate system due to changes in solar forcing differ from changes due to CO2, and is the response sensitive to the sign of the forcing?

  • Model: CGCM

  • Length of integration: 300 yrs (150 yrs each)

  • Lead PIs: Chris Bretherton, Roger Marchangd, and Bjorn Stevens

Abrupt 2x and 0.5xCO2 (NonlinMIP)

  • Experiment names: abrupt-2xCO2, abrupt-0p5xCO2

  • Science question: To what extent is regional-scale climate change per CO2 doubling state-dependent (nonlinear); what are the associated mechanisms; and how does this affect our understanding of climate model uncertainty?

  • Model: CGCM

  • Length of integration: 300 yrs (150 yrs each)

  • Lead PIs: Peter Good

AMIP with preindustrial forcing 1870-present

  • Experiment names: amip-piForcing

  • Science question: Are climate feedbacks during the 20th century different to those acting on long term climate change?

  • Model: AGCM

  • Length of integration: 145 yrs

  • Lead PIs: Tim Andrews

Time slice experiments forced with CO2 and SSTs from AMIP, piControl, and abrupt 4xCO2 runs

  • Experiment names: piSST (preindustrial SST), piSST-4xCO2, piSST-4xCO2-rad, piSST-pxK, a4SST, a4SSTice, a4SSTice-4xCO2, amip-a4SST-4xCO2

  • Science question: How do regional climate responses (e.g. in precipitation) and their uncertainties in coupled models arise from the combination of different aspects of CO2 forcing and sea surface warming?

  • Model: AGCM

  • Length of integration: 307 yrs

  • Lead PIs: Rob Chadwick, Herve Douville, and Chris Skinner

Diagnostic packages


  • COSP simulator outputs for various instruments (ISCCP, CALIPSO, PARASOL, CloudSat, MODIS, MISR) are requested in DECK AMIP

  • A lighter set of ISCCP and CALIPSO simulator outputs are also requested to interpret and assess the credibility of cloud feedbacks (monthly, daily) in DECK + CMIP6 historical, CFMIP Tier 1, Abrupt +/- 4% solar forcing, AMIP -4K, COOKIE LW off

  • Lead PIs: Steve Klein and Alejandro Bodas-Salcedo

  • See COSP page for more information.

Instanteneous high-fequency outputs at cfSites locations

  • 3-hourly instanteneous outputs at several locations are requested in amip, amip-p4K, and amip-4xCO2

  • St. Helena Eastern Tropical Ocean (Paquita Zuidema), DYNAMO & ARM MJO Experiments (Simon deSzoeke) (Optional), ARM site location for GoAmazon2014/5 (David Neelin) (Optional), and EUREC4A locations? (Sandrine Bony, Bjorn Stevens TBC)

  • Lead PIs: Mark Webb

The CFMIP/CMIP6 experimental protocol is now fully described in the CFMIP3/CMIP6 paper in a GMD Special Issue on CMIP6.

Webb, M. J., Andrews, T., Bodas-Salcedo, A., Bony, S., Bretherton, C. S., Chadwick, R., Chepfer, H., Douville, H., Good, P., Kay, J. E., Klein, S. A., Marchand, R., Medeiros, B., Siebesma, A. P., Skinner, C. B., Stevens, B., Tselioudis, G., Tsushima, Y., and Watanabe, M.: The Cloud Feedback Model Intercomparison Project (CFMIP) contribution to CMIP6, Geosci. Model Dev., 10, 359-384, doi:10.5194/gmd-10-359-2017, 2017.

The original CFMIP proposal for CMIP6 was endorsed by the CMIP Panel in September 2015 and minor updates to experiment names, descriptions and lengths were made in Feb 2016.

For details of the CFMIP experiments and diagnostics proposed for CMIP6 please refer to the CFMIP/CMIP6 experiment and diagnostic specification (Updated Feb 26th 2016). This information is currently being incorporated into the Central CMIP6 Data Request.

Please raise any issues with mark.webb (at)