access coupled model for cmip5 ipcc ar5
play

ACCESS Coupled Model for CMIP5/IPCC AR5 Daohua (Dave) Bi 1 and - PowerPoint PPT Presentation

ACCESS Coupled Model for CMIP5/IPCC AR5 Daohua (Dave) Bi 1 and Martin Dix 1 , Maciej Golebiewski 1 , Tony Hirst 1 , Simon Marsland 1 , Siobhan OFarrell 1 , Petteri Uotila 1 , Arnold Sullivan 1 , Zhian Sun 2 , Xiaobing Zhou 2 , Ian Watterson 1


  1. ACCESS Coupled Model for CMIP5/IPCC AR5 Daohua (Dave) Bi 1 and Martin Dix 1 , Maciej Golebiewski 1 , Tony Hirst 1 , Simon Marsland 1 , Siobhan O’Farrell 1 , Petteri Uotila 1 , Arnold Sullivan 1 , Zhian Sun 2 , Xiaobing Zhou 2 , Ian Watterson 1 Centre for Australian Weather and Climate Research (CAWCR) A partnership between CSIRO and the Bureau of Meteorology 1 CSIRO Atmospheric and Marine Research, Aspendale 2 Bureau of Meteorology, Melbourne CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  2. Outline 1. Introduction to the ACCESS coupled model Framework • 2. Performance of ACCESS coupled model in simulating world climate, especially the world ocean climate Overview: progress in ACCESS coupled model development • Assessment of a few recent runs: Sea Surface Temperature • Convection, Overturning and Barotropic Streamfunction • Model skill scores in simulating a set of key climatic fields, globally • and over Australia Most recent progress • 3. Ongoing work (instead of conclusion) CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  3. ACCESS/AusCOM Coupled Model Framework (configuration for IPCC AR5/CMIP5) � Atmospheric model UM Atmosphere � Resolution: 192 x 145 1.875 ° lon x 1.25 ° lat UM 7.3 Land Surface 38 levels in the vertical MOSES or CABLE 1.4 Coupler OASIS 3.2.5 Ocean Sea Ice CICE 4.1 MOM4p1 Ocean Biogeochemistry AusCOM1.0 release Users Guide (Bi, D. and S. Marsland): http://www.cawcr.gov.au/publications/technicalreports/CTR_027.pdf CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  4. ACCESS/AusCOM Ocean-Sea Ice Tripolar Grid Horizontal: Global tripolar with resolution of 360 x 300; Longitudinal: uniform 1°; Latitudinal: Equatorial meridional refinement: 1/3° 10°S-10°N; Mercator grid in Southern Ocean: 1° at 30°S to 0.25° at 78°S. MOM4 and CICE share this horizontal grid. Vertical: 50-level ocean covering 0-6000m with a resolution ranging from 10m (for above 200m column) to 250 m (for the abyssal ocean) CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  5. ACCESS/AusCOM Coupling Strategy UM/MATM i2a a2i cpl cpl CICE i2o o2i cpl cpl cpl MOM4 •Different coupling frequencies for atm � ice (3 hours), and ice � ocean (every time step, ie, ½ or 1 hr) •CICE functions as a ‘coupling buffer’/’media’ between UM and MOM •62 coupling fields (2D): a2i 24, i2o 13, o2i 7, i2a 18 CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  6. 2. ACCESS Climate Simulations: Basic Assessment (Test control runs) ACCESS coupled model has very short history: “assembled” successfully in mid 2009, fully functional since early 2010, and has since then been going through extensive debugging, tuning and re-configuring processes (“slow” but steady progress): • UM HadGEM3 setup: Tens of century scale “present” climate control simulations have been performed over the period from February 2010 to present. • UM HadGEM2 setup: a couple of 100-year runs done recently, under debugging... • Why TWO versions? Criteria for assessment (suitable for AR5/CMIP5?): • Size of global average SAT or SST (regional) bias and (global) drift. • Overall skill in simulating a set of key climatic fields, globally and over Australia. • Realism of simulation of the mean state in the tropical Indo/Pacific Ocean region. • Realism of simulation of ENSO and influence on Australian rainfall. • Realism of polar region sea ice extent. • Realism of world ocean circulations, especially SO, including strength of the Antarctic Circumpolar Current and extent of late winter deep convection. • Extent of any other substantial biases evident. • …… CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  7. Overview of progress in ACCESS coupled model development in the past 12 months (hg3 setup): ‘warming up’ test1: 46-level ocean; Cooling test2: 50-level ocean; test3: 6hr � 3hr A � I coupling; > 3 °C test4: reduction in ice albedo testx: one of the “cold” runs; testy: cloud inhomogeneous scaling (overtuning) testz: with “proper” cloud scaling parameter CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  8. A Few More Recent Runs (hg3 and hg2 setup for UM and various ocean/ice configurations, mostly 200 yrs): Global ocean T Global ocean S Global SSS Global SST Base-09 Base-10 Base-11 Hg2-5A Hg2-6A C-000 CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  9. Global Ocean Overturning Circulations Base-09 Base-10 Base-11 Hg2-5A Hg2-6A C-000 CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  10. Barotropic Streamfunction and ACC Transport Through Drake Passage Sv (10 6 m 3 s -1 ) CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  11. Ann Max Mixed Layer Depth 10-year Average hg3 setup runs hg2 setup runs CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  12. SST Biases hg3 setup runs hg2 setup runs CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  13. ENSO Spectrum CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  14. ACCESS Simulation of key climatic fields (global) (M-skill test for one of the ACCESS hg3 runs) Ta: SAT; SWg: downward SW at ground; LW ↑ & SW ↑ : upward LW & SW at TOA ; Cf: Total cloud radiative forcing; Ct: total cloud cover The mean skill scores across all 9 (global) fields are 0.695, 0.736 and 0.785; Over Australia, scores for these fields are 0.629, 0.685 and 0.814. CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  15. The 1st CABLE Run (hg3 setup, with no cloud tuning): Very Promising CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  16. CABLE Run: Hope and Issues MOSES run SST bias CABLE – MOSES CABLE – MOSES (yrs 11~20 avg) (yrs 11~20 avg) (yrs 59~68 avg) Global MOC yrs 59~68 avg Global SST evolution MOSES run CABLE run CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  17. 3. Summary and ongoing work • ACCESS coupled model is still under development (extensive testing and tuning still under way); • Progress is ‘slow’ but steady, and recent improvement associated with CABLE implementation shows good potential (but with “bug” to be fixed in CABLE); • Code and configuration will be frozen by late June 2011, based on the best performance achieved by then; • ‘Formal’ AR5/CMIP5 runs will be performed from July 2011 (in parallel); • Wish for good luck… CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  18. Dr Daohua Bi ACCESS Coupled Modeller Centre for Australian Weather and Climate Research CSIRO MAR Aspendale Vic 3199 Email: dave.bi@csiro.au Web: www.accessimulator.org.au Thank you. CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  19. IPCC AR5 timelines Analysis papers for IPCC must be IPCC WG1 submitted by final plenary 31 July 2012 2010 2011 2012 2013 2014 Model output Data after this point CMIP5 continues to will probably not be starts becoming accept model results included in AR5-cited available to users well after IPCC AR5, at publications via the ESG least through 2013 CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  20. NH Sea Ice Coverage (Base-12c/Base-12b) NH ice area (10 6 km 2 ) Base-12c Base-12b CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  21. SH Sea Ice Coverage (Base-12c/Base-12b) SH ice area (10 6 km 2 ) Base-12c Base-12b CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  22. ACCESS Series/Parallel Coupling CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  23. Impact of ITF transport on the T.P. cooling Note Base-02B is implemented with proper Rayleigh drag scheme Which is meant to reduce the water mass (and thus heat) transport From T. Pacific to Indian Ocean through the Lombok Strait, and Therefore reduces the TP cooling bias... Annual Mean SST Diff: Base-02B – Base-02A (years 101-120) CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

  24. Ice Coverage from HadGEM2 runs Hg2_15m4 Hg2_15m5 N.H S.H. CSIRO. GREENHOUSE 2011, Cairns, April 4-8, 2011

Recommend


More recommend