Update on CAM2.X Development and Future Research Directions

1
Update on CAM2.X Developmentand Future Research
Directions

• J. J. Hack
• National Center for Atmospheric Research
• Boulder, Colorado USA

On behalf of Phil Rasch and Leo Donner, CCSM
AMWG Co-Chairs
2
Outline

• Model biases in CAM2
• Physics changes in CAM 2.X
• Simulation improvements in CAM 2.X
• Future directions

3
Principal CAM2/CCSM2 Biases

• Warm winter land surface temperature bias
• Cold tropical tropopause temperatures
• Double ITCZ and extended cold tongue
• Erroneous cloud response to SST changes
• Deficiencies in E. Pacific surface energy budget
• Underestimation of tropical variability

All work over last year aimed at reducing these
biases
4
CAM 2.X Physics Changes Relative to CAM2

• Moist Physics and Clouds
• improved prognostic cloud water moist processes
• tighter interaction of shallow convection and
  cloud water
• transfer of mixed phase precipitation to land
  surface
• improved cloud parameterization
• Radiation
• shortwave forcing by diagnostic aerosols
• updated SW scheme for H2O absorption
• updated LW scheme for LW absorption and emission
• Modeling Extensions
• reintroduction of Slab Ocean Model (SOM)
• Other
• energy fixers for dynamical cores plus related
  diagnostics
• additional diagnostic capabilities
• updated boundary datasets
• implementation improvements
• high-resolution configuration

5
Examples of Simulation Improvements
NH Winter Land Surface Temperatures
6
Examples of Simulation Improvements
Tropopause Temperatures
7
Examples of Simulation Improvements
Shortwave Response to ENSO
CAM2
CAM2.X
ERBE (ERBS)
8
Notable Simulation Changes
Surface Insolation Changes (primarily aerosol
effects)
9
High-resolution configurations

• T85 spectral Eulerian configuration
• workhorse for IPCC
• 2x2.5 Finite-Volume configuration
• in process
• High-resolution simulation improvements
• warmer tropospheric temperatures
• improvements in low-level circulation

10
T85 Zonal Annual Mean Temperatures
CAM2.X T85
CAM2.X T42
CAM2.X T85-CAM2.X T42
11
T85 vs T42 Cloud Forcing
T85 vs T42
12
T85 Surface Wind Stress
T85
T42
Diff
13
Current Assessment of High-Resolution Coupled
Simulation (oceanographers perspective)

• reductions in warm biases off the western coasts
  of the continents
• reductions in southern ocean SST errors
• improvements in the near equatorial upper-ocean
  temperature structures
• improved semi-annual signal of equatorial Pacific
  SSTs
• improved Pacific equatorial undercurrent
  (increased westward wind stress)
• improved surface salinity changes in the Arctic
• poleward shift of southern hemisphere storm track

14
Future Directions

• Return to science-driven investment
• understanding, improving, and evaluating key
  processes in physical climate system
• attack problems from a fundamental perspective
• holistic focus on phenomenology that deals with
  nonlinear interactions of processes
• accept incremental improvements based on
  fundamental process advancements
• focus on developing a better understanding of
  simulation performance
• evaluation of climate sensitivity, feedbacks,
  forcing, etc.
• identify and fill holes in existing expertise for
  dedicated attention
• boundary layer, cloud/aerosol microphysics,
  numerical methods
• Coordinate research activities on physical
  climate system with activities that extend
  simulation capabilities
• examples include aerosol, chemical, and
  biogeochemical modeling extensions

15
Future Directions (continued)

• Extensions to modeling capabilities
• Aerosol modeling
• enhance/strengthen focused effort
• aerosol chemistry, links to cloud microphysics
  (e.g., indirect effect), atmospheric chemistry
• constrain process models and associated physical
  parameterization development
• Atmospheric chemistry
• enhance/strengthen ongoing efforts
• e.g., interactions with ACD and external
  collaborators
• revived CCSM working group?
• quantify requirements and constraints on
  development of the physical climate system
• Integration of offline transport modeling
  capability
• replace MOZART and MATCH

16
Future Directions (continued)

• Extensions to modeling capabilities
• Middle-Atmosphere and Climate (middle
  atmospheric dynamics/physics)
• continued development of WACCM
• Upper Troposphere Lower Stratosphere (UTLS)
  modeling initiative
• emphasis on water transport/interactions
• immediate driver on development of improved
  numerical approximations
• Biogeochemistry
• support and leverage division and external
  efforts on carbon cycle modeling
• stress treatment of boundary-layer and
  convective-scale transport mechanisms
• e.g., mineral dust
• leverage links to atmospheric chemical modeling
• interactive surface processes including chemical
  interactions

17
Future Directions (continued)

• Where would we like to be in 5 years?
• highly-integrated physical parameterization
  package
• boundary layer - moist convection - stratiform
  cloud processes
• explore tighter linkages available w/ regard to
  radiation and clouds?
• significantly enhanced large-scale dynamical
  driver(s)
• 2-3X increase in default horizontal resolution in
  each dimension
• improved vertical resolution as dictated by
  physics, dynamics, or numerics
• e.g., boundary layer, resolution of tropopause
  height
• capability for local resolution refinement
• incorporation of adaptive gridding techniques?
• formally conservative transport capabilities
• isentropic formulation(s)?
• isotropic discretizations in spherical geometry?
• non-hydrostatic formulations??
• is this essential to scientific work on the
  proposed time scale?

18
Future Directions (continued)

• Where would we like to be in 5 years?
• fully-interactive aerosol modeling capabilities
• links to cloud microphysics to accommodate work
  on indirect effect
• fully interactive atmospheric chemical modeling
  capability
• stratospheric and tropospheric formulations
• necessary hooks/linkages to biogeochemical cycles
  
• be creating hierarchy of modeling tools along the
  way
• simplified column physics frameworks through
  fully-coupled system model
• provides a powerful diagnostic and evaluation
  framework for understanding
• be exploiting opportunities w/ regard to
  assimilation capabilities
• CAPT parameterization evaluation framework
• development of relationships with NASA NCEP,
  others?

19
Summary

• CAM2.X model driven by need to reduce some major
  systematic biases
• Incorporates major changes to parameterized
  physics
• Incorporates extensions to modeling and
  diagnostic capabilities
• High-resolution configuration shows promising
  behavior
• Strong foundation to build upon

20
END
21
Other Notable Simulation Features
Large differences in cloud amount
CAM2
CAM2.X
CAM2.X-CAM2
22
Extension to Include Convective Cloud Fraction

• Some improvements to cloud distribution and
  associated SWCF
• Clear improvements in short wave radiative
  response to ENSO

23
Other Notable Simulation Features
Large differences in cloud condensate
CAM2
CAM2.X
CAM2.X-CAM2
24
2nd backup
25
Example of extensions to cloud scheme

• Majority of CAM clouds are diagnosed as a
  function of relative humidity
• stratocumulus cloud coverage diagnosed from
  Klein-Hartmann stability metric
• generally limited to small areas in the eastern
  portion of ocean basins
• remainder of cloud relative humidity dependent
• moisture biases contributed to bi-modal vertical
  distribution of cloud

26
Convective Cloud Fraction

• Introduce cloud diagnostic based on local
  convective mass flux
• cloud fraction logarithmic function of cloud mass
  flux

27
Convective Cloud Fraction
28
Convective Cloud Fraction

• Some improvements to cloud distribution and
  associated SWCF
• Clear improvements in short wave radiative
  response to ENSO

29
Second change motivated by problems with physics
package at high resolution and in FV dynamical
core

• Inability to maintain extratropical cloud forcing
  at high resolution
• Similar problems with cloud forcing using
  finite-volume dynamical core

T42
T42
T85
T85
30
Standard Physics Tuning Implications
Implied ocean heat transport
T85 vs T42
31
Deficiencies in extratropical cloud forcing
fundamentally related to deficiencies in cloud
fraction scheme (relative humidity clouds)
32
Provide additional source of cloud liquid water
from shallow convection process

• Shallow convection scheme designed to deal with
  shallow and mid-level instabilities
• philosophical framework based on redistribution
  of water, as opposed to rainout
• Applied as cleanup procedure following
  application of deep convection

33
Detrainment of cloud liquid water to prognostic
clouds from shallow convection

• Explicitly calculate the minimum rainwater
  autoconversion
• drizzle rates in trade cumulus regimes little
  change to behavior
• 2-3 mm/day rainfall rates in deep convective
  regimes
• Detrain remainder of required condensation to
  cloud water scheme
• provides sufficient additional degree of freedom
  allowing scaling of cloud scheme
• enables a more portable physics package across
  dynamical cores

34
Enhanced Physics Cloud Forcing (intermediate
result)
T85 vs T42
35
Current Assessment of High-Resolution Coupled
Simulation (oceanographers perspective)

• reductions in the warm biases off the western
  coasts of the continents
• enhanced upwelling in the eastern oceans
• reductions in southern ocean SST errors
• improvements in the near equatorial upper-ocean
  temperature structures
• improved semi-annual signal of equatorial Pacific
  SSTs
• improved Pacific equatorial undercurrent due to
  increased westward wind stress
• improved surface salinity changes in the Arctic
• poleward shift of southern hemisphere storm track