CFMIP-CMIP6 Experiments

Transcription

1 CFMIP-CMIP6 Experiments Mark Webb, Timothy Andrews, Alejandro Bodas-Salcedo, Sandrine Bony, Chris Bretherton, Robin Chadwick, Hélène Chepfer, Hervé Douville, Peter Good, Jennifer Kay, Stephen Klein, Roger Marchand, Brian Medeiros, Pier Siebesma, Chris Skinner, Bjorn Stevens, George Tselioudis, Yoko Tsushima and Masahiro Watanabe. CFMIP Meeting, Tokyo, September 2017

2 The Cloud Feedback Model Inter-comparison Project (CFMIP) Understanding and evaluating clouds, circulation and precipitation in present and future climates. CFMIP Committee George Tselioudis, Masahiro Watanabe (Co-chairs) Timothy Andrews, Sandrine Bony, Chris Bretherton, Yen-Ting Hwang, Sarah Kang, Jen Kay, Steve Klein, Thorsten Mauritsen, Mark Webb

3 Cloud Feedback Model Inter-comparison Project Objective 1: Inform improved assessments of climate change cloud feedbacks by: a) improving our understanding of cloud-climate feedback mechanisms. b) improving evaluation of clouds and cloud feedbacks in climate models. CFMIP3/CMIP6 Hierarchy T/q budgets, cfsites CFMIP Intercomparison of SCM/LES Exploitation of Sat Obs via Simulators in CMIP CMIP6/CFMIP3 CGILS COSP Crown copyright Met Office IPCC AR6 WCRP Assessment

4 Cloud Feedback Model Inter-comparison Project 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. Precipitation impact of PBL radiative effects: Fermepin and Bony 2014 Precipitation response to CO 2 and SST pattern changes: Chadwick 2016 Crown copyright Met Office Effect of second doubling of CO2 on precipitation over Africa: Good et al (NonLinMIP)

5 Three Categories of GCM Experiments Relevant to CFMIP Phase 3 CMIP DECK + CMIP6 Historical Experiments - AMIP, preindustrial control, 1% CO2, abrupt 4xCO2, CMIP6 Historical CFMIP/COSP PMC have proposed COSP + process diagnostics for these. CFMIP experiments in CMIP6 Described in Webb et al., GMD, 2017 Modelling groups participate via CMIP6 activity on CMIP6 timescales (to 2020) Hosted on the Earth System Grid as part of CMIP6 activity Required Tier 1 experiments and optional Tier 2 experiments Participation now confirmed for 15 models: HadGEM3-GC31-LL, MIROC6, MRI- ESM2, BCC-CSM2-MR, CESM2, MPI-ESM1.2-LR, GFDL-CM4, TaiESM1, NICAM.16, CNRM-CM6, CIESM, CanESM5, GISS-E2.1, IPSL-CM6, NorESM2. All Tier 1 experiments expected to be pulished by end 2018, Tier 2 by mid CFMIP-3 informal experiments (next talk)

6 CFMIP CMIP6 Experiment Summary Pre-industrial clouds and precipitation Historical / present day clouds and precipitation CO 2 forcing, cloud and precipitation adjustments Climate feedbacks and precipitation responses picontrol CMIP6 historical abrupt-4xco2 1pctCO2 CMIP/CMIP6 amip amip-4xco2 amip-p4k amip-future4k CFMIP Tier 1 CFMIP Tier 2 pisst aqua-control aqua-4xco2 aqua-p4k abrupt-solp4p abrupt-solm4p amip-piforcing abrupt-2xco2 amip-m4k abrupt-0p5xco2 amip-lwoff amip-p4k-lwoff aqua-control-lwoff aqua-p4k-lwoff pisst-4xco2-rad a4sst pisst-pxk pisst-4xco2 a4sstice a4sstice-4xco2 amip-a4sst-4xco2 Clouds Circulation and Precipitation

7 CFMIP Cloud Observation Simulator Package (COSP) COSP simulator outputs for various instruments (ISCCP, CALIPSO, PARASOL, CloudSat, MODIS, MISR) are requested in AMIP DECK A lighter set of ISCCP and CALIPSO simulator outputs are also requested to interpret and assess the credibility of cloud feedbacks (monthly, daily) DECK + CMIP6 Historical, CFMIP Tier 1, Abrupt +/- 4% Solar, Amip -4K, COOKIE LW off PI s Steve Klein and Alejandro Bodas-Salcedo. Too many papers published using COSP outputs from CMIP5 to list! See CFMIP publications page on CFMIP website. Much analysis of CMIP6 COSP outputs anticipated. Crown copyright Met Office

8 Use of Temperature and Humidity Tendency terms to understand cloud feedback mechanisms (e.g. Brient et al 2015) Also Webb and Lock, 2013; Demoto et al., 2013; Yoshimori et al 2013,2014; Sherwood et al., 2014; Kamae and Watanabe, 2012; Ogura et al., 2014; Brient et al., in prep. Yukari Takayabu: Convective latent heating observed from space: From TRMM to GPM CMIP DECK, CFMIP Tier 1, +/-4% solar, COOKIE LW off Crown copyright Met Office

9 Instantaneous high frequency outputs at cfsites locations (PI Mark Webb) St. Helena Eastern Tropical Ocean (Paquita Zuidema) DYNAMO & ARM MJO Experiments (Simon deszoeke) (Optional) ARM site location for GoAmazon2014/5 (David Neelin) (Optional) EUREC4A locations? (Sandrine Bony, Bjorn Stevens TBC) cfsites will be requested in amip, amip-p4k and amip-4xco2 Plans to develop approaches based on existing work by Webb, Neggers, Dal Gesso, Nuijens, Mülmenstädt, Nygren, Roehrig Crown copyright Met Office

10 Tier 1 CFMIP-CMIP6 Experiments 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? Tier 1 experiments retain CFMIP-2/CMIP5 idealized experimental hierarchy (amip, amip-p4k, amip-4xco2, amip-future4k, aqua-control, aqua-4xco2 and aqua-p4k: 174 yrs atmosphere only AGCM) Lead coordinator: Mark Webb Crown copyright Met Office Ringer et al, 2015.

11 Tier 2: AMIP minus 4K Experiment (PIs Mark Webb, Bjorn Stevens) Science Question: Are cloud feedbacks consistent for climate cooling and warming, and, if not, why? HadGEM2-A +4K response 0.27 Wm -2 K -1 Yoshimori et al. (2009) found a positive shortwave cloud feedback in a future CO2 MIROC3 experiment but a negative feedback in cooling PMIP LGM experiments. Confirmed in pilot amip +/- 4K experiments HadGEM2-A, experiments without ice sheets or SST pattern differences. Cloud feedback assumed symmetric in most climate sensitivity estimates from palaeoclimate data - but may be out by a factor of 2! How important in other models? Why? HadGEM2-A -4K response 0.14 Wm -2 K -1 Clausius Clapeyron water vapour gradient asymmetry? Overturning response asymmetry? More cloud ice at lower sunnier latitudes? Adiabatic cloud liquid water contents? Plan: Test hypotheses using tendency terms + HadGEM2- A sensitivity tests (e.g. linearize Clausius-Clapeyron as in Joshi et al., 2008)

12 Tier 2: amip-piforcing: feedbacks in response to observed SST variations AGCMs forced with observed monthly SST and sea-ice variations, i.e. an extended amip experiment, but with forcings held constant at picontrol levels. GCMs show decadal variations in feedback, mostly from clouds, due to changing SST patterns. Feedback is always larger (climate sensitivity smaller) than that found in 4xCO2 experiments. Gregory, J.M., and T. Andrews, 2016: Variation in climate sensitivity and feedback parameters during the historical period. Geophys. Res. Lett., 43, , doi: /2016gl

13 Tier 2: amip-piforcing: feedbacks in response to observed SST variations AGCMs forced with observed monthly SST and sea-ice variations, i.e. an extended amip experiment, but with forcings held constant at picontrol levels. Feedback parameter 4xCO 2 vs amip Pilot multi-model study reveals robust decadal variations in climate feedback parameter in response to changing 20 th century SST patterns, as well as model diversity. In all models feedback is larger (climate sensitivity smaller, ~2K) in response to historical SSTs than that in 4xCO 2 experiments (2-4.5K) with implications for observed estimate of climate sensitivity. Timothy Andrews et al., in preparation.

14 Atmosphere-only experiments without LW cloud-radiative effects (Tier 2) Sandrine Bony and Bjorn Stevens Science Question: How do cloud-radiative effects impact the structure, the strength and the variability of the general circulation in present and future climates? Rationale: An increasing number of single model studies shows evidence for the critical impact of cloud-radiative effects on the large-scale circulation (ITCZ, extra-tropical eddies), its variability (MJO, ENSO) and its organization (convective aggregation). How robust are these results? In the past, we promoted COOKIE (Stevens et al. 2012), i.e. experiments in which atmospheric cloud-radiative effects (ACRE) are removed both in the LW and in the SW. But these experiments are associated with significant changes in land surface temperature and land-sea contrasts Surface temperature changes between clouds-off and clouds-on (Stevens et al. 2012)

15 Atmosphere-only experiments without LW cloud-radiative effects (Tier 2) Sandrine Bony and Bjorn Stevens Science Question: How do cloud-radiative effects impact the structure, the strength and the variability of the general circulation in present and future climates? Rationale: An increasing number of single model studies shows evidence for the critical impact of cloud-radiative effects on the large-scale circulation (ITCZ, extra-tropical eddies), its variability (MJO, ENSO) and its organization (convective aggregation). How robust are these results? To better isolate the impact of atmospheric cloud-radiative effects on the circulation, we propose LW-COOKIE: clouds are made transparent to LW radiation only after Webb et al. GMD (2017) Analysis plan: still open & to be discussed; we will help coordinate one or several collective papers; different people might take the lead on the analysis of specific aspects of the circulation response. Contact: Sandrine Bony & Bjorn Stevens

16 Tier 2 : +/- 4% Abrupt solar forcing AOGCM experiments : abruptsolp4p and abrupt-solm4p PI s: Chris Bretherton, Roj Marchand, and Bjorn Stevens Background: Solar forcing differs from greenhouse forcing because of differences in the vertical structure of warming, which impacts atmospheric stability, clouds and precipitation. A +4% solar experiment abrupt-solp4p is proposed which is analogous to the abrupt-4xco2 experiment but rather than changing CO2 it would abruptly increase the solar constant by four percent and keep it fixed for 150 years, resulting in a global mean radiative forcing of a similar magnitude to that due to CO2 quadrupling. The complementary -4% abrupt solar forcing experiment will allow the examination of feedback symmetry under climate cooling, and would also help with the interpretation of model responses to volcanic forcing and geoengineering scenarios (e.g. GeoMIP G1) where CO2 and solar forcing are simultaneously changed. Plans: Examine regional cloud and precipitation changes across models, including fast and slow responses. Requests: Please generate at least ISCCP simulator output. COSP simulated CALIPSO cloud fractions and MISR histograms are also desired (and computationally cheap).

17 Tier 2: Timeslice experiments for understanding regional climate responses to CO 2 forcing. Co-ordinators: Rob Chadwick (Met Office), Hervé Douville (CNRM) and Chris Skinner (U. Michigan) Seven 30 year AGCM timeslice experiments, designed to decompose coupled model regional climate responses to CO 2 forcing, and one 36 year amip-type experiment for examining the impact of SST biases. Science questions: How do regional climate responses in a coupled model arise from the combination of responses to different aspects of CO 2 forcing and warming (uniform SST warming, pattern SST warming, sea-ice change, direct CO 2 effect, plant physiological effect)? Which aspects of forcing/warming are most important for causing inter-model uncertainty in regional climate projections? Can inter-model differences in regional projections be related to underlying structural or resolution differences between models through improved process understanding, and could this help us to constrain the range of regional projections? What impact do coupled model SST biases have on regional climate projections?

18 Drivers of DJF Atlantic storm-track changes DJF Circulation Change HadGEM2 Crown copyright Met Office Chadwick et al. 2017, Clim. Dyn.

19 Tier 2: abruptco2 experiments for quantifying and understanding state-dependent regional climate responses to CO 2 forcing (nonlinmip). Co-ordinator: Peter Good (Met Office) Two experiments: abrupt2xco2 and abrupt0.5xco2 (as CMIP6 abrupt4xco2 but with different CO2 concentration. Science aims: quantifying and understanding climate responses: 1. Under forcing ~comparable to that under Paris agreement (abrupt2xco2 control) 2. Climate change avoided by mitigation (abrupt4xco2 abrupt2xco2). 3. Non-linear change (the difference between 2 and 1) 4. Change under past cold climates (abrupt0.5xco2 control) 5. Non-linear change with larger signal/noise (abrupt4xco2- abrupt2xco2) (control abrupt0.5xco2)

20 State-dependent precipitation responses to CO 2 in 5 GCMs Precipitation change over Africa is state dependent (from nonlinmip pilot study) Doubling difference (% of control): Second CO2 doubling minus first CO2 doubling First doubling: ~rcp4.5 at 2100 abrupt2xco2-control Second doubling: mitigation benefits abrupt4xco2-abrupt2xco2 Good et al., 2012: A step-response approach for predicting and understanding nonlinear precipitation changes. Clim Dyn Chadwick, R. and Good, P., Understanding non-linear tropical precipitation responses to CO2 forcing. GRL Good, Peter, et al. "Nonlinear regional warming with increasing CO2 concentrations." Nature Climate Change 5.2 (2015): Bouttes, N., et al. "Nonlinearity of ocean heat uptake during warming and cooling in the FAMOUS climate model." Geophysical Research Letters 42.7 (2015): Crown copyright Met Office

21 Crown copyright Met Office

22 CFMIP CMIP6 Experiments (Tier 2) Are cloud feedbacks consistent for climate warming and cooling, and if not, why? - AMIP minus uniform 4K SST (amip-m4k 36yrs AGCM) - Mark Webb and Bjorn Stevens (Link to PMIP) How do cloud-radiative effects impact the structure, the strength and the variability of the general atmospheric circulation in present and future climates? - Atmosphere-only experiments with clouds transparent to longwave radiation (amip-lwoff, amip-p4k-lwoff, aqua-control-lwoff, aqua-p4k-lwoff: 92 yrs AGCM) - Sandrine Bony and Bjorn Stevens How do responses in the climate system due to changes in solar forcing differ from changes due to CO 2, and is the response sensitive to the sign of the forcing? - Abrupt +/-4% Solar Forcing (abrupt-solp4p, abrupt-solm4p: 300 yrs AOGCM) - Chris Bretherton, Roger Marchand, Bjorn Stevens To what extent is regional-scale climate change per CO 2 doubling state-dependent (nonlinear); what are the associated mechanisms; and how does this affect our understanding of climate model uncertainty? - Abrupt 2x and 0.5x CO 2 (abrupt-2xco2,abrupt-0p5xco2 300yrs AOGCM) - Peter Good (NonLinMIP) Crown copyright Met Office

23 CFMIP CMIP6 Experiments (Tier 2) Are climate feedbacks during the 20 th century different to those acting on long term climate change? - AMIP with preindustrial forcing 1870-present (amip-piforcing 145yrs AGCM) - Timothy Andrews 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? -Timeslice experiments forced with CO 2 and SSTs from amip, preindustrial and abrupt4xco2 - preindustrial SST, pi SST + uniform SST, pi SST + 4xCO2-rad, pi SST + 4xCO2 - a4 (abrupt 4CO2) SST, a4 SST+sea ice, a4 SST+sea ice+4xco2 - amip + a4 SST response + 4xCO2 - (307 yrs AGCM) - Robin Chadwick, Hervé Douville, Chris Skinner Crown copyright Met Office

24 Cloud Feedback Model Inter-comparison Project (CFMIP) contribution to CMIP6. Crown copyright Met Office

25 CFMIP CMIP6 Experiment Summary Pre-industrial clouds and precipitation Historical / present day clouds and precipitation CO 2 forcing, cloud and precipitation adjustments Climate feedbacks and precipitation responses picontrol CMIP6 historical abrupt-4xco2 1pctCO2 CMIP/CMIP6 amip amip-4xco2 amip-p4k amip-future4k CFMIP Tier 1 CFMIP Tier 2 pisst aqua-control aqua-4xco2 aqua-p4k abrupt-solp4p abrupt-solm4p amip-piforcing abrupt-2xco2 amip-m4k abrupt-0p5xco2 amip-lwoff amip-p4k-lwoff aqua-control-lwoff aqua-p4k-lwoff pisst-4xco2-rad a4sst pisst-pxk pisst-4xco2 a4sstice a4sstice-4xco2 amip-a4sst-4xco2 Clouds Circulation and Precipitation

26 CFMIP CMIP6 Experiment Summary Pre-industrial clouds and precipitation Historical / present day clouds and precipitation CO 2 forcing, cloud and precipitation adjustments Climate feedbacks and precipitation responses picontrol CMIP6 historical abrupt-4xco2 1pctCO2 CMIP/CMIP6 amip amip-4xco2 amip-p4k amip-future4k CFMIP Tier 1 CFMIP Tier 2 pisst aqua-control aqua-4xco2 aqua-p4k abrupt-solp4p abrupt-solm4p amip-piforcing abrupt-2xco2 amip-m4k abrupt-0p5xco2 amip-lwoff amip-p4k-lwoff aqua-control-lwoff aqua-p4k-lwoff pisst-4xco2-rad a4sst pisst-pxk pisst-4xco2 a4sstice a4sstice-4xco2 amip-a4sst-4xco2 Clouds Circulation and Precipitation

27 CFMIP CMIP6 Experiment Summary Historical / present day clouds and precipitation CO2 forcing, cloud and precipitation adjustments Climate feedbacks and precipitation responses picontrol CMIP6 historical CMIP/CMIP6 amip amip-4xco2 amip-p4k CFMIP Tier 1 aqua-control aqua-4xco2 aqua-p4k abrupt-4xco2 1pctCO2 amip-future4k abrupt-solp4p abrupt-solm4p amip-piforcing abrupt-2xco2 amip-m4k abrupt-0p5xco2 CFMIP Tier 2 pisst amip-lwoff amip-p4k-lwoff aqua-control-lwoff aqua-p4k-lwoff pisst-4xco2-rad a4sst pisst-4xco2 a4sstice a4sstice-4xco2 amip-a4sst-4xco2 pisst-pxk Circulation and Precipitation Clouds Pre-industrial clouds and precipitation

28 CFMIP CMIP6 Experiment Summary Historical / present day clouds and precipitation CO2 forcing, cloud and precipitation adjustments Climate feedbacks and precipitation responses picontrol CMIP6 historical CMIP/CMIP6 amip amip-4xco2 amip-p4k CFMIP Tier 1 aqua-control aqua-4xco2 aqua-p4k abrupt-4xco2 1pctCO2 amip-future4k abrupt-solp4p abrupt-solm4p amip-piforcing abrupt-2xco2 amip-m4k abrupt-0p5xco2 CFMIP Tier 2 pisst amip-lwoff amip-p4k-lwoff aqua-control-lwoff aqua-p4k-lwoff pisst-4xco2-rad a4sst pisst-4xco2 a4sstice a4sstice-4xco2 amip-a4sst-4xco2 pisst-pxk Circulation and Precipitation Clouds Pre-industrial clouds and precipitation

29 Difficulty in interpreting physical feedback mechanisms in coupled model experiments due to lack of process diagnostics Temperature and humidity tendency terms have been demonstrably useful for understanding the roles of different parts of the model physics in cloud feedbacks and adjustments Williams et al 2013, Webb and Lock 2013, Kamae and Watanabe 2012, Demoto et al 2013, Sherwood et al 2014, Ogura et al They have also been used to understand regional warming patterns such as polar amplification in coupled models (e.g. Yoshimori et al 2013,2014) Ogura et al 2014 Yoshimori et al 2014