1. Representation of Subgrid Cloud-Radiation Interaction and its Impact on Global Climate Simulations Xinzhong Liang (Illinois State Water Survey, UIUC ) Sunwook Park and Liping Deng (ISU) Xiaoqing Wu Department of Geological and Atmospheric Sciences Iowa State University Partly by DOE CCPP and ARM

2. 1. The problem 2. Parameterization of subgrid cloud-radiation interaction by mosaic approach 3. Impacts on global climate simulations

3. CCSM3 1. The problem

4. Radiation Parameterization Scheme in General Circulation Model (GCM) Radiative transfer equations for shortwave and longwave fluxes and heating rates Representation of cloud optical properties such as cloud emissivity and optical depth using cloud liquid and ice water paths Treatment of cloud horizontal inhomogeneity and vertical overlap Cloud Geometry and Inhomogeneity General Circulation Model

5. 2. Parameterization of subgrid cloud-radiation interaction by mosaic approach and evaluation against CRM simulations Liang and Wang (1997, JGR) Wu and Moncrieff (2001, JAS) Liang and Wu (2005, GRL)

6. C 1-C GCM MOS Mosaic approach (MOS) of treating subgrid cloud variability ( Liang and Wang 1997, JGR )

7. Convection Radiation Clouds CRM

8. Cloud liquid/ice water mixing ratio (g/kg) CRM approach GCM approach Wu and Moncrieff (2001, JAS) Quantifying cloud variability effects

9. Cloud Base Height Cloud Top Height CRM Cloud Frequency Cc Ci Cs CRM simulated cloud frequency (10 -4 ) distribution as a function of the base and top heights. Four major cloud clusters are identified in the centers as deep convective tower (Cc), anvil cirrus (Ci), and stratiform cloud (Cs) that are distinguished by the mosaic treatment of cloud-radiation interactions. Evaluating mosaic approach against CRM simulations Wu, Parker and Liang (2005) Liang and Wu (2005, GRL)

10. Procedures for evaluating mosaic approach (MOS) 1)Search for all unbroken segments of cloud layers where total cloud water path ( ) is larger than 0.2 g m -2. 2)GCM-grid mean cloud liquid and ice water profiles ( ) are obtained by averaging over the entire domain and scaling with total cloud fraction. 3)If a segment has its base below 4.5 km and top above 9.5 km, it is grouped into the Cc cluster; otherwise if its base is above 9.5 km and at the base, then the segment falls into the Ci cluster; all the remaining ones belong to the Cs cluster. 4)GCM-grid mean cloud cover fractions for Cc, Ci and Cs are determined from all respective segments over the whole domain. 5)Mosaic treatment with subcells is then performed once using the CRM-based GCM-grid mean profiles scaling by. Liang and Wu (2005, GRL)

11. Shortwave Flux TOA Surface CRM GCM CRM MOS CRM GCM CRM MOS Liang and Wu (2005, GRL)

12. Longwave Flux TOA CRM GCM CRM MOS CRM GCM CRM MOS Surface Liang and Wu (2005, GRL)

13. SWLWTotal Domain average shortwave (SW), longwave (LW) and total heating rate (K/day) profiles as simulated by the CRM and calculated by the GCM and mosaic (MOS) approaches. Liang and Wu (2005, GRL) Radiative Heating Rate

14. 3. Impacts on global climate simulations Wu and Liang (2005, GRL)

15. In-cloud water concentration derived from the CRM simulation (solid) compares with that used by the standard CCM3 (dashed). Circles are normalized CRM cloud (liquid and ice) water paths. Wu and Liang (2005, GRL)

16. High-level Cloud (%)Total Cloud Liquid Water Path (g/m 2 ) GCM MOS ISCCP GCM MOS SSM/I Wu and Liang (2005, GRL)

17. 5-year (79-83) global averages of radiative fluxes (W/m 2 ) from observations (OBS), CCM3 (GCM) and mosaic run (MOS) TOA SRF

18. GCM MOS-GCM Radiative Heating Rate (K/day) Wu and Liang (2005, GRL)

19. Temperature (K) GCMGCM-NCEPMOS-GCM Wu and Liang (2005, GRL)

20. Summary The inclusion of subgrid cloud-radiation interaction through mosaic approach in the radiation scheme of GCM enables the use of more realistic cloud amounts and cloud water contents while producing net radiative fluxes closer to observations. The inclusion of subgrid cloud-radiation interaction through mosaic approach in the radiation scheme of GCM enables the use of more realistic cloud amounts and cloud water contents while producing net radiative fluxes closer to observations. Consequently, not only the representation of cloud- radiation interactions is more physically consistent and accurate, but also climate simulations are affected and improved. Consequently, not only the representation of cloud- radiation interactions is more physically consistent and accurate, but also climate simulations are affected and improved.

21. MOS-CCM3 (1979-1983) ANN LW SW TOA

22. MOS-CCM3 (1979-1983) ANN LW SW Surface

23. In-cloud water concentration derived from the CRM simulation (solid) compares with that used by the standard CCM3 (dashed). Circles are normalized CRM cloud (liquid and ice) water paths. Wu and Liang (2005, GRL)

24. Flux(Wm -2 ) OBSMOS/ocCTL/mc F LW (TOA)233.9245.6225.7 F LW (SRF)49.471.252.2 F SW (TOA)234.0251.2213.1 F SW (SRF)165.9186.8145.8