Australian CMIP6 Modelling Plans ACCESS-CM2/ESM2 Simon Marsland CSIRO Oceans and Atmosphere WGCM-19, Dubrovnik, October 20, 2015

ACCESS – Australian Community Climate and Earth System Simulator ACCESS-CM2 (AOGCM for CMIP6) Atmosphere – UK Met Office GA7.0 N96, L85, UM10.x, pending Ocean – NOAA/GFDL MOM5.1 Sea ice – LANL CICE5.1 Land surface – CABLE2.0 Community Atmosphere Biosphere Land Exchange, pending ACCESS-ESM2 (ESM for CMIP6) ACCESS-CM2+ Terrestrial biogeochemistry - CASA-CNP Oceanic biogeochemistry – WOMBAT (Matear, CSIRO) Simon Marsland, WGCM_19, Dubrovnik, October 2015

ACCESS-CM2 Resolution Standard Resolution High Resolution • Atmospheric resolution – N96 (~1.2°lat; ~1.8°lon); 85 levels • Oceanic resolution – ~1 deg. (enhanced tropics, high latitudes) • Prototype configured – runs at 5 years per day • Will form basis of ACCESS-ESM2 High Resolution • Atmospheric resolution – N216 (~0.55°lat; ~0.8°lon), 85 levels, not computat. viable • Oceanic resolution – 0.25 degrees horizontal for DECK, historical, ScenarioMIP • Subject to adequate computational resources • AOGCM only Simon Marsland, WGCM_19, Dubrovnik, October 2015

CMIP6 Timeline ACCESS-ESM2 (N96, 1 degree) • Perform extensive set for CMIP6 ~2016/17/18 • ~8000 years of simulation for CMIP6 ACCESS-CM2 (N96, 0.25 degree) • Perform limited set ~2017/18 • ~2000 years of simulation • spin-up + DECK + historical + limited ScenarioMIP Contact: Simon Marsland (CSIRO) + ACCESS MIP leads Simon Marsland, WGCM_19, Dubrovnik, October 2015

Indian Ocean Dipole (SON) versus ocean resolution Years 101-200 ACCESS-CM2 1 degree HadISST (yrs 1950-2014) ACCESS-CM2 ¼ degree IOD strong influence on Australian temperature and rainfall patterns Moderates (-ve counteracts/+ve enforces) El Nino influence on Australia Simon Marsland, WGCM_19, Dubrovnik, October 2015

ACCESS CMIP6 Participation Why? Demonstrate ACCESS is a world-class system CMIP6 developed to address important science questions (GC’s etc.) Large uptake via Earth System Grid CMIP6 Model Intercomparison Project commitments DECK – Tony Hirst (CSIRO) Scenario MIP (revised ‘RCP’s) – Tony Hirst CFMIP (cloud feedback) – Charmaine Franklin (CSIRO) DAMIP (detection/attribution) – David Karoly (ARCCSS) C4MIP, LS3MIP, LUMIP (land use) – Rachel Law (CSIRO) OMIP, FAFMIP (oceans) – Simon Marsland (CSIRO) GeoMIP (geoengineering) – Andrew Lenton (CSIRO) Simon Marsland, WGCM_19, Dubrovnik, October 2015

Summary and future work Work to do UM10.x with GA7 CABLE2.x implementation Scalability performance improvement Model tuning - scientific performance (about a year) Risks CMIP6 funding for ACCESS-CM2 is available (NESP) Funding for ACCESS-ESM2 less certain (but likely, CSIRO) Compute/Storage: NCI commitment to CMIP6? Model untested: Fallbacks ACCESS1.4/ESM1 and ACCESS-CM2 (UM GA6.0) ACCESS staffing has declined 25% over last 3 years, but more experienced and growing collaboration with ARCCSS (5x Universities in Centre of Excellence) Simon Marsland, WGCM_19, Dubrovnik, October 2015

CMIP5 model inter-comparison: M-Score based on model skills in representing 3 key atmospheric fields: SAT, MSLP, and Precipitation Global Australia Simon Marsland, WGCM_19, Dubrovnik, October 2015

ACCESS-ESM1 simulations Prescribed atmospheric CO2 Land carbon exchange (5 yr smoothing) 1000 yr pre-industrial simulations completed. Long spin-up for carbon cycle. (Blue) Historical simulations (green) Future scenarios: RCP4.5 (black) RCP8.5 (red) Ocean carbon exchange Comments For the highlights (e.g. cloud highlight) can there be more or a different description? If not on slide, verbally would be good. Last year: Initial ACCESS-CM2 configured and (technical) test simulation performed Powerful new model evaluation methodology implemented ACCESS1.0/1.3 delivery to CMIP5 continued ACCESS1.4 model version completed Carbon cycle implemented to form ACCESS-ESM1 University uptake of ACCESS and CABLE Ocean uptake increases as CO2 increases, Land initially uptake then source to atmosphere

Issues Land water/ice to ocean Preindustrial Volcanic forcing Icesheets and Glaciers (mid-latitude) Still no systematic way of treating this in CMIP historical or Scenarios Implications for regional signature of sea-level rise related to steric changes Plans for CMIPn+1? Preindustrial Volcanic forcing Gregory et al. GRL, 2013 Climate models without piControl volcanic forcing underestimate historical ocean thermal expansion Ocean strong cooling jump in 1850 historical, effect lasts century(s) WGCM recommendation Simon Marsland, WGCM_19, Dubrovnik, October 2015

Thank you Simon Marsland Ocean and Coupled Modelling Team Leader Earth System Assessment Program t +61 3 9239 4548 e simon.marsland@csiro.au w http://people.csiro.au/M/S/Simon-Marsland csiro oceanS and atmosphere