2/27/06CAB Meeting Scientific and Technical Plans for IPCC AR5 Bill Collins With input from Lawrence Buja and Jerry Meehl National Center for Atmospheric.

2/27/06 CAB Meeting

Scientific and Technical Plans for IPCC AR5

Bill CollinsWith input from Lawrence Buja and Jerry Meehl

National Center for Atmospheric Research

Boulder, Colorado

2/27/06 CAB Meeting

Our principal scientific targets(CAB Meeting, June 2005)

• Major objective: Develop, characterize, and understand the most realistic and comprehensive model of the observed climate system possible.

• Subsidiary objectives:– Analyze and reduce the principal biases in our

physical climate simulations using state-of-the-art theory and observations.

– Simulate the observed climate record with as much fidelity as possible.

– Simulate the interaction of chemistry, biogeochemistry, and climate with a focus on climate forcing and feedbacks.

2/27/06 CAB Meeting

Interaction of the CCSM community with the IPCC process

• We design the model we want for Earth system science -- this model will naturally be ideal for IPCC.

• Our collective objective should be novel experiments with CCSM4 --these add tremendous value above and beyond the IPCC scenarios.

2/27/06 CAB Meeting

Possible components of CCSM4(CCSM Plenary, June 2005)

Atmosphere Ocean

Coupler

Sea IceLand

C/NCycle

Dyn.Veg.

Ecosystem & BGCGas chem.Prognostic

AerosolsUpperAtm.

2/27/06 CAB Meeting

Configuration of CCSM4: Open issues

• The physical components:– What improvements will be ready to address long-standing biases?– What major modifications will be included in CAM?– What resolutions will we use? What can we afford?

• The chemical and biogeochemical components:– What terrestrial ecosystem model will we use?– What mechanism for gaseous chemistry will we use?– What representation of aerosols will be adopted?– What lid will we adopt for the upper atmosphere?– Will dynamic vegetation be compatible with the

simulated climate?

• Computational cost:– What changes can we make to the numerical formulation to

increase scaling by 10x? By 25x?

2/27/06 CAB Meeting

The timeline for IPCC AR5(CAB Meeting, June 2005)

Assumptions:• It is likely that the AR5 report will be issued 6 years after AR4, in 2013.

• Following the precedent in AR4, the simulations will have to be finished two to three years ahead, in mid-2011 at the latest.

• Therefore we need to have CCSM4 ready in 2009 (at the latest).

• This means we have just four years to accomplish our objectives.

2006 2007 2008 2009 2010 2011 2012 20132005

AR5 ProcessCCSM4 Development CCSM4 Release

2/27/06 CAB Meeting

A more detailed timeline for IPCC runs

Lawrence Buja

2/27/06 CAB Meeting

Lessons Learned from IPCC AR4(CAB Meeting, January 2005)

1. Start with an experimental plan

2. Develop forcings early in the process of experimental design and model development

3. Design and test methods for equilibration of CCSM with present-day and pre-industrial conditions.

4. Plan and test data processing to insure timely data transfer to IPCC archive(s).

5. Develop scientific requirements for:a. Model capabilities: e.g., chemical processes, etc.

b. Spatial resolution of the CCSM component models

c. The size of the ensembles for 20th C. and scenario runs

2/27/06 CAB Meeting

Goals of the design exercise• Identify major scientific objectives of our IPCC runs

– What do we want to explore using 20c3m simulations?

– What hypotheses are we testing with scenario runs?

– What can we learn about our climate model and the climate from the multi-model ensemble?

• Identify subsidiary scientific objectives: for example– Tests of new methods for understanding climate sensitivity and

climate feedbacks.

– Development of state-of-art emissions data sets and simulations of forcing agents (e.g., BC and OC aerosols)

• Map/integrate these objectives into the CCSM science plan• Design the simulations:

– Spin-up process

– Ensemble size

• Create an implementation plan

2/27/06 CAB Meeting

Proposal for draft outline of IPCC plan

1. Introduction: the interaction of CCSM and IPCCA. What have we learned from the 20c3m runs?

B. What have we learned from multi-model papers?

C. What’s new in the IPCC AR5?

2. Scientific objectives for the IPCC AR5A. Major objectives: e.g., quantifying carbon-cycle feedbacks

B. Minor objectives: e.g. quantifying aerosol indirect effects

3. Description of the forcing agents:A. Well-mixed GHGs and the C/N cycle

B. Tropospheric and stratospheric ozone

C. Aerosols: direct, semi-direct, and indirect effects

D. Solar variation

E. Land-use change: effects of agriculture, fires, urbanization

2/27/06 CAB Meeting

Proposed outline (cont.)

4. Description of the experiments:A. Requirements for ensemble size

B. Experiments to determine model resolution

C. Description of 20c3m and scenario integrations

D. Subsidiary experiments, e.g., pre-industrial controls, etc.

5. Description of the model spin-up procedure:A. Initial conditions for prognostic fields

B. Acceleration techniques (if any)

C. Evolution or boundary conditions for forcing agents

6. Implications for the CCSM program:A. Coordination on model configuration and capabilities

B. Coordination on forcing agents

C. Coordination on configuration of experimental configuration.

2/27/06 CAB Meeting

Proposed outline (cont.)

7. Resource requirements:A. Computer time:

1) Estimation of CPU hours / year of simulation2) Estimation of total CPU time required for IPCC runs and model

development leading to IPCC version of CCSM4.3) Plan for acquisition of computer time.

B. Human resources:1) Identification of the major software engineering projects.2) Estimation of the person-years of SE effort required.

C. Post-processing, archival, and data transfer1) Estimate of model data per simulation year2) Total data volume for the IPCC simulations3) Identification of archival sites 4) Estimation of processing and transfer throughput

8. Timelines and a decision tree9. Distribution and analysis of IPCC simulations

2/27/06 CAB Meeting

Planning by WGCM and other groups

• Workshop in summer 2006 for modeling groups and scenario developers for IPCC AR5:– Participants include:

• WGCM• AIMES• SPARC• Other IGBP groups

– Product of meeting: recommendations for AR5 scenarios

• Meeting of WGCM and AIMES in September 2006:– Product: white paper outlining plans for ESMs & scenarios

– CCSM should present its plans and configuration for AR5 at this meeting.

2/27/06 CAB Meeting

Provisional list of AR5 required simulations

• Standard runs from all groups:– Equilibrium mixed-layer 2xCO2 runs

– 1%/year 2xCO2 run for TCR from AOGCMs

– All-forcings integrations for 20th century

– All-forcings integration for 21st century with 1 scenario (A1B?)

• New adaptation/mitigation scenario integrations for ESMs

2/27/06 CAB Meeting

Conclusions

• Provisional timeline for CCSM4 and IPCC AR5:– Freeze of model formulation in late 2008

– Public release in summer 2009

– Initiation of historical simulations in spring 2010

– Completion of scenario integrations in spring 2011

• This schedule gives three years for model development.

• We should consider a coordinated process to resolve open issues for the configuration of CCSM4.

• This process should stem from our experimental objectivesfor studies of coupled Earth system.

2/27/06 CAB Meeting

Some big chemistry questions

• What do we want chemistry for?– Short-lived forcing agents:

• Aerosols

• Ozone

• How much of the chemistry must be interactive?• What’s the lid on the model?

– Stratopause– Mesopause– Higher?

• Where do we get the historical emissions data sets for 30-90 chemical species?

• What are your metrics for fidelity?