1. CMIP6 Catherine Senior C Senior, WGRC Montevideo, 14th March 2014
MOHC Seminar, December 3rd, 2013
C Senior, WGRC Montevideo, 14th March 2014
© Crown copyright Met Office

2. CMIP history
The Coupled Model Intercomparison Project (CMIP) is organised by the World Climate Research Programme (WCRP) Working group on Coupled Models (WGCM)
current WGCM co-chairs are Sandrine Bony and Catherine Senior (just taken over from Jerry Meehl)
Within WGCM there is a CMIP panel (chaired by Veronika Eyring) that directly oversees CMIP activities
CMIP provides a coordinated multi-model experiment framework to address compelling science questions and new scenarios
CMIP is not dictated by IPCC
CMIP5 was the first distributed data source and a significant step-up in terms of coordinated experimental design
© Crown copyright Met Office

3. CMIP5 experience
CMIP5 experimental design (Taylor et al 2009) has created an enormous resource and enabled a large amount of science
Large numbers of idealised experiments
Complex RCP process but reasonable scenarios generated
Emissions v Concentrations simulations
First effort at near term prediction
Core v tiers
Distributed data sources
Earth System Grid Federation (ESGF) used for data archival and retrieval
Lots of problems/delays in governance and delivering workable solution
Metadata
Lack of stable funding
© Crown copyright Met Office

4. Development, Evaluation and Characterization (DECK) Experiments
CMIP6: Towards understanding past, present and future climate
CMIP 6 will be a distributed organization which will;
Establish a set of CMIP ongoing model development evaluation and characterization experiments to gain basic information about model performance and sensitivity;
done by most groups as part of the development cycle.
revisited whenever a new model was developed
The basis for the Model Intercomparison Projects
Evolve only slowly (10-15 yr time scales)
CMIP Panel continues to manage the details of these experiments
Provide a benchmarking and evaluation tool run on ESGF
Development, Evaluation and Characterization (DECK) Experiments
AMIP (~ ), Pre-industrial control, 1%/yr CO2 to quadrupling, Instantaneous 4xCO2, Historical simulation-> RCP8.5 (concentrations for AOGCMs, emissions/concentrations for ESMs?)
© Crown copyright Met Office

5. CMIP6: Towards understanding past, present and future climate
CMIP 6 will be a distributed organization which will;
Promote sets of experiments to address science questions within the context of the WCRP Grand Challenges and AIMES input specific to CMIP6:
Systematic biases
Response to forcings
Variability, predictability and future scenarios
WCRP Grand Challenges
Clouds, Circulation and Climate Sensitivity
Changes in Cryosphere
Climate Extremes
Regional Climate Information
Regional Sea-level Rise
Water Availability
Theme for collaboration: biospheric forcings and feedbacks
© Crown copyright Met Office

6. CMIP6: Towards understanding past, present and future climate
CMIP 6 will be a distributed organization which will;
Around these experiments build CMIP6 with additional, specialized intercomparisons (“MIPs”) that would make use of the same standards and infrastructure.
Individual MIPs work with CMIP Panel to coordinate details of experimental design and variable lists and etc.; each MIP would designate which experiments would contribute to the CMIP6 science questions and thus target for wider participation of many modeling groups, and they could have other specialized experiments for their own communities
CMIP Panel has oversight and identifies the elements of the MIP experiments that are part of CMIP6
MIPs can run throughout CMIP6 cycle not necessarily all at end
MIPs
CFMIP, PMIP, CORDEX, Transpose-AMIP, C4MIP, CCMI, Decadal (DCPP), GeoMIP, …
LUMIP, ScenarioMIP
© Crown copyright Met Office

7. Systematic Bias Response to forcings Variability, predictabiity,
Ocean
processes
Other MIP experiments
Cloud
feedbacks
Model
errors
AMIP
CMIP6
experiments
Systematic Bias
Chemistry/
stratosphere
Paleo-
climate
CMIP:
Ongoing model
development,evaluation,
characterization
standardization
coordination,
infrastructure,
documentation
Scenarios
Emission
engineering
Geo-
Response to forcings
Variability,
predictabiity,
future scenarios
Decadal
prediction
Carbon
cycle
Detection/
attribution
Regional
climate/
extremes
Land use 7

8. Communication and responsibilities
CMIP Panel (5 people) facilitates communication between MIP co-chairs and the model group contacts to help with coordination between MIPs, and between the MIPs and the modeling groups. Assisted by WGCM members
CMIP Panel:
Coordinate diagnosis and evaluation simulations with the community
Approve experiments and variable lists etc. that are to be part of CMIP6
Coordinate with WCRP Grand Challenges
MIPs:
Address WCRP Grand Challenges and science questions
Suggest model simulations to address these science questions
Work with CMIP Panel for output list for CMIP6 data request, and supply rationalization for relevance of experiments to address science questions
MIPs determine which experiments are run when
© Crown copyright Met Office

9. Infrastructure
Earth System Grid Federation (ESGF) provides the backbone for the increasingly integrated system for distributing model (and observational) output
Ability to use the software effectively critically depends on development and maintenance of standards for file formats, metadata, cataloguing, publication, node management etc – huge challenge as delegation of effort to MIPs increases
Urgent need to find a route for modelling groups to participate in increasing numbers of MIPs without required logistical effort becoming untenable.
Proposed WGCM Infrastructure Panel (WIP) which will consist of those responsible for infrastructure and also users e.g. those from modelling centres managing data provision and retrieval
Establish and maintain standards for MIPs
Possibly seek funding to provide properly tested standardised software
© Crown copyright Met Office

10. Systematic Bias Response to forcings Variability, predictabiity,
Ocean
processes
Other MIP experiments
Cloud
feedbacks
Model
errors
AMIP
CMIP6
experiments
Systematic Bias
Chemistry/
stratosphere
Paleo-
climate
CMIP:
Ongoing model
development,evaluation,
characterization
standardization
coordination,
infrastructure,
documentation
Scenarios
Emission
engineering
Geo-
Response to forcings
Variability,
predictabiity,
future scenarios
Decadal
prediction
Carbon
cycle
Detection/
attribution
Regional
climate/
extremes
Land use 10

11. CMIP6: next steps Lots of upcoming workshops and meetings
Plans for CMIP6 to be published in EOS and the open for comment
Final design will be agreed at WGCM18 in October 2014
© Crown copyright Met Office

12. HadGEM3-GC2 experiments
Plan to perform a small sub-set of runs at N216 (60km) and N96 (135km) from CMIP5:
AMIP (3.3)
Pre-industrial control (3.1) to 500 years
Historical ( ) (3.2) 2?x156 years
RCP 8.5 ( ) (4.1) 95 years
1% year-1 to 4xCO2 (6.1) 140 years
Step change to 4xCO2 (6.3) 150 years
Emphasis initially on idealised scenarios rather than realistic RCPs. Understand impact of resolution (TCR, ECS, mechanisms of regional climate change) rather than focus on new projections.

13. ScenarioMIP
A pair of two new scenarios (high/low) run by all models with new 20th century (to 2015) and 21st century forcings
Scenario MIP matrix, modeling groups that subscribe run a
pair of two additional new scenarios (Sample different combinations of scenario pairs and AOGCMs/ESMs (sampled in an appropriate way, e.g. climate sensitivity, enough realizations)
Paired non-mitigation/mitigation scenarios
Scenario
Pair 1
Scenario
Pair 2
Scenario
Pair 3
Scenario
Pair 4
Scenario
Pair 1
IAM and climate modeling community decides which scenario pairs make most sense:
baseline/mitigation scenario pairs for research on benefits of mitigation related to land use change, short lived climate forcers, etc.
An overshoot scenario
Model 1
AOGCMs and ESMs x
Model 2 x
new emissions and concentrations need to be generated) 3 x x 4 .

16. Summary
CMIP6 will be a distributed organisation with much more engagement from MIPs in designing experiments and responsibility for diagnostics
Science focus will be around WCRP Grand Challenges
Challenge for data service and delivery through ESGF. Communication and application of standards will be critical to success
New scenarios for near and long term and strong links to Climate Services
HadGEM2-GC2 is our first step in ‘continuous model development’ and we are essentially running DECK experiments now
© Crown copyright Met Office

17. Radiative feedbacks
© Crown copyright Met Office