1. Evaluation of a scheme representing cloud inhomogeneous structure in the Australian Community Climate and Earth System Simulator (ACCESS)
Zhian Sun, Xiaobing Zhou, Yimin Ma, Hongyan Zhu (Bureau),
Charmaine Franklin,, Dave Bi, Martine Dix, Tony Hirst, Simon. J. Marsland, Maciej Golebiewski (CSIRO),
Cyril Morcrette (Met Office),
John Shonk (Reading University)

2. Outline Introduction to ACCESS coupled climate model
Tripleclouds scheme: cloud horizontal inhomogeneity in grid-box
Exponential–random overlap scheme: Vertical structure of clouds
Experimental results
Summary
The Centre for Australian Weather and Climate Research A partnership between CSIRO and the Bureau of Meteorology

3. ACCESS component models and systems
Puri (2005) project plan
OPS
Initial priorities:
NWP for Bureau
operations
Coupled climate
modelling for IPCC AR5 &CMIP5
NWP
Assimilation
4DVAR
Dynamic Vegetation
/Disturbance
Atmosphere UM
Assimilation
Coupler
OASIS
Land surface
CABLE
Atmos. chemistry
UKCA
Terrestrial
C,N,P cycles
Ocean
GFDL MOM4
Sea ice
CICE
Oceanic
biogeochemistry
Assimilation
BODAS
OBS

4. MOM4P1: Atmosphere: (1) Horizontal point : 360x300.
Longitudinal direction: uniform 1°, Latitudinal direction is 1/3° from oS to 10oN. Mercator grid is used in this region with 1 degree at 30oS to
0.25 degree at 78oS.
(2) Vertical direction: 50 levels covering m with a 10m-resolution
from the surface to 200m.
(3) A global tripolar grid point.
Atmosphere:
HadGEM 2/3 atmospheric models
N96 – 1.875 lon x 1.25  lat; 38 levels

5. Final result for IPCC AR5
Initial result in 2009
Final result for IPCC AR5
SST bias from ACCESS coupled model
The Centre for Australian Weather and Climate Research A partnership between CSIRO and the Bureau of Meteorology

6. Why original model has cold SST bias?
Energy from atmosphere may not be large enough to heat the ocean
Surface wind may not be right for driving the ocean surface current
The Centre for Australian Weather and Climate Research A partnership between CSIRO and the Bureau of Meteorology

7. Net SW flux at surface
The Centre for Australian Weather and Climate Research A partnership between CSIRO and the Bureau of Meteorology

8. Problem is likely related to cloud, radiation
Clouds are very complex in structure and interact strongly with radiation.
Hence they are a very important aspect of the climate system and need suitable representation in climate models.
However, representation of cloud in climate model is simple due to model resolution and computational limitation.
The Centre for Australian Weather and Climate Research A partnership between CSIRO and the Bureau of Meteorology

9. The Cloud Modelling Challenge
We make two main approximations when representing clouds in a radiation code.
Horizontal structure: optical depth in a gridbox layer is described by a single value of cloud water/ice content.
Vertical structure: cloud is aligned vertically according to an overlap assumption.
Cloud in a gridbox layer can then be represented by just two quantities: a cloud fraction and a single value of LWC/IWC.

10. The Cloud Modelling Challenge
Assumption of homogeneous cloud distribution overestimates effect of cloud on radiation, leading to cooling at the surface
Visible satellite image from 08 November 2010 at 12:00. 10
The Centre for Australian Weather and Climate Research A partnership between CSIRO and the Bureau of Meteorology

11. Tasks to improve representation of cloud
Cloud fraction
We modified cloud scheme to improve cloud amount prediction (not present here)
Cloud horizontal inhomogeneous
Implement an inhomogeneous cloud scheme developed by Shock at Reading University
Vertical overlap
Implement vertical exponential-random overlap scheme by Hogan & Illingworth from Reading University

12. Schematic demonstration of triplecloud
Homogeneous cloud
Tripleclouds
Mean cloud water content
Standard deviation of
The Centre for Australian Weather and Climate Research A partnership between CSIRO and the Bureau of Meteorology
The Centre for Australian Weather and Climate Research A partnership between CSIRO and the Bureau of Meteorology

13. The Tripleclouds method
With Tripleclouds, the continuous water content distribution in a gridbox layer is represented with a two-point distribution function.
water content
Continuous distribution
frequency
Plane-parallel: one value
Tripleclouds: two values
standard deviation
Fractional standard deviation fw = 0.75

14. Cloud overlap: random + maximum
The Centre for Australian Weather and Climate Research A partnership between CSIRO and the Bureau of Meteorology

15. From Hogan & Illingworth (QJ 2005)
max
ran
Overlap parameter 
Using radar observations to work out cloud cover of pairs of levels as function of level separation and values calculated using different overlap assumptions
The Centre for Australian Weather and Climate Research A partnership between CSIRO and the Bureau of Meteorology

16. Z0 decorrelation distance
z is level separation
Z0 decorrelation distance
Shock analysed radar data in different locations
and found Z0 has a latitude dependence
The Centre for Australian Weather and Climate Research A partnership between CSIRO and the Bureau of Meteorology

17. Net SW radiation at surface
EXP1:Inhomo-cloud
EXP2:cloud fraction
EXP4: exponential-random overlap

18. SST biases

19. SST bias from ACCESS coupled model
The Centre for Australian Weather and Climate Research A partnership between CSIRO and the Bureau of Meteorology

20. Summary
SST in ACCESS coupled-model simulation has been significantly improved
The Improvement is mainly due to the modifications in the UM model physics
1) Horizontal inhomogeneous representation of triplecloud scheme
2) New vertical exponential-random overlap scheme
3) Modification to the cloud fraction for both water and ice clouds
4) Modification to the surface wind roughness length
All these changes apply to ACCESS and have been used in IPCC AR5 & CMIP5 experiments

21. Thank you Zhian Sun Phone: 96694408 Email: z.sun@bom.gov.au
The Centre for Australian Weather and Climate Research
A partnership between CSIRO and the Bureau of Meteorology
Zhian Sun
Phone:
Web:
Thank you