the community climate system model

CAS2K3 Meeting, AnnecySeptember 8-11, 2003


The Community Climate System The Community Climate System ModelModel

Bill CollinsChair, CCSM Scientific Steering Committee

National Center for Atmospheric ResearchBoulder, Colorado

• Applications of CCSM• Development of new version of CCSM• Features of CCSM simulations• Future improvements in CCSM• Computational demands for new science



Structure of CCSM

Atmosphere(CAM)

Ocean(POP)

Coupler(CPL)

Sea Ice(CSIM)

Land(CLM)



Component Models in CCSM

• Atmosphere– Multiple dynamical cores: SLD, Eulerian, & Finite Volume– Generalized 2D decomposition of grid– Resolutions with most heritage: T31, T42, and T85 (L26)

• Ocean– Derivative of LANL Parallel Ocean Program – Grid: spherical in S. hemisphere, orthogonal curvilinear in N.

hemisphere– Standard resolution: 320 x 384 (L40)

• Land Surface– Superset of NCAR LSM and Georgia Tech BATS– Same horizontal resolution as atmosphere– 10 layers for soil, up to 5 for snowpack

• Sea Ice– Up to 5 categories of sea-ice thickness



Organization of CCSM Project



Utilization of CCSM



Simulation of Recent Climate



CCSM2 – Holocene Climate

CCSM2

Kohfield and Harrison, 2000

Annual Precipitation Change8.5 ka minus Present

mm/year

• PMIP models underestimated both the north-ward shift and magnitude of the precipitation increase required to maintain lakes and vegetation. • CCSM2 more realistically simulates rainfall increases between 20-25°N latitude• CCSM2 still underestimates needed rainfall along the Mediterranean and in South Africa.

Collaborators: PARCS, PMIP

For the mid-Holocene, lake levels, plant macrofossil, and pollen data indicate a greener Sahara.



CCSM2 – Last Interglacial Climate

Dye 3

Renland

Camp Century

GRIP

CCSM2

• CCSM2 forced with anomalous seasonal solar insolation associated with orbital changes warms the Arctic climate.

• For summer, Arctic sea ice decreases significantly and Greenland warms by 1-5oC.

• Terrestrial data sites warmed rapidly and early; most by 4-6oC.

• Greenland ice cap melted except for summit.

JJA Surface Temperature130 ka minus Present



CCSM2 Improvements for Paleoclimate

• River runoff scheme• New sea-ice model• Free surface ocean &

prognostic cloud water in the atmosphere

• Grid and ocean/land mapping capability for any past time period

Bathymetry and land mask editing tool eCubed. [John Davis and Rick Smith (LLNL); Walter Voit (UT Dallas)]



El Nino Simulations for Current Climate



Predicting Co-evolution of CO2 and Climate

Atmospheric GCMPrognostic CO2

Ocean GCM+ BGC

Biophysics+ BGC

Fossil FuelCO2



Ongoing Development of CCSM

CSM 1.0

CCSM 2.0

New ocean, land, sea-ice models New physics in atmosphere

June 1996

May 2002

CCSM 3.0

New physics in all models

Winter 2003



Main Model Biases in CCSM2

Double ITCZ and extended cold tongue Cold tongue mitigated in new POP, double ITCZ in new CAM

Overestimation of winter land surface temperatures Bias largely eliminated in new CLM and CAM

Underestimation of tropical tropopause temperatures Bias reduced in new CAM

Erroneous cloud response to SST changes Signs and pattern of response corrected in new CAM

• Errors in E. Pacific surface energy budget Overestimation of insolation larger in prototype

• Underestimation of tropical variability Underestimation still present



Double ITCZ



Winter Land Surface Temperatures



Tropical Tropopause Temperatures



Cloud Forcing Response to Tropical SSTs



East Pacific Surface Energy Budget



Tropical Variability



Ocean Model Changes for CCSM 3.0

• KPP boundary layer • Double Diffusion• Solar Absorption (Chlorophyll)• Ideal Age and Passive Tracer

Infrastructure• Ocean Currents in Air-Sea Fluxes



CONTROL All Physics + Numerical MODS

Jan

July

Mixed Layer Depths



SST Signature of Solar Absorption



SST Signature of All Ocean Modifications



Goal For CCSM2: state-of-the-art model with focus on biogeophysics, hydrology, river routing

Goal For Next Few Years: Natural and human-mediated changes in land cover and ecosystem functions and their effects on climate, water resources, and biogeochemistry

Process: Continue to develop and improve existing physical parameterizations in the model while adding new biological and chemical process

Land Model Working Group

Hydrology

Drainage

Canopy Water

Evaporation

Interception

SnowMelt

Sublimation

ThroughfallStemflow

Infiltration Surface Runoff

Evaporation

Transpiration

Precipitation

Soil Water

Redistribution

Direct Solar

Radiation

Absorbed SolarRadiation

Diff

use

Sol

ar

Rad

iatio

n

Long

wav

e R

adia

tion

Reflected Solar Radiation

Em

itted

Lon

g-w

ave

Rad

iatio

n

Sen

sibl

e H

eat

Flu

x

Late

nt H

eat

Flu

x

ua0

Momentum FluxWind Speed

Soil Heat Flux

Heat Transfer

Pho

tosy

nthe

sis

Biogeophysics



Biogeochemistry

Goal: Enable parameterization of terrestrial carbon cycle

• New hierarchical data structures within a grid cell: grid cell land unit snow/soil columns plant functional types. Does not change climate.• Allow multiple plant types on a single soil column. Minor climate changes.

Biogeophysics I

Goal: Minor changes or bug fixes

• Allow precipitation to be rain/snow mix • Improve 2-m temperature diagnostic • Remove discontinuity in bulk density of snow

Biogeophysics II

Goal: Reduce excessive summertime temperatures in arid regions. Reduce Arctic winter warm temperature bias

• Turbulent exchange of heat from ground

Model Development for new CLM



Increased Turbulent Heat Flux from Vegetated SurfacesCLM2.2 CLM2



Effects on Land Surface Temperatures

• 22-year (1979-2000) simulations of CAM2 with CLM2.2• CLM2.2 significantly cools surface air temperature

CLM2.2 CLM2



Changes to Physics in CAM3

• Clouds and condensate:Clouds and condensate:– Improved prognostic cloud water & moist processes – Transfer of mixed phase precipitation to land surface– Improved cloud parameterization

• Radiation:Radiation:– Shortwave forcing by diagnostic aerosols – Updated SW scheme for H2O absorption– Updated LW scheme for LW absorption and emission

• Surface models:Surface models:– Introduction of CLM 2.2– Reintroduction of Slab Ocean Model (SOM)

• Energy fixers for dynamics + diagnosticsEnergy fixers for dynamics + diagnostics



Increased Cloud Condensate

• Separate cloud liquid and ice variables• Advect cloud condensate• Include latent heat of fusion• Use ice & water variables for cloud optics• New dependence on temperature for cirrus particle size• Sedimentation of cloud droplets and ice particles• Modified evaporation of rain

CAM 2CAM 3

CAM 3 – CAM 2



Increased Cloud Amounts

• PBL height constrained• Rain rate > 0• Convection cloud amounts from convective mass fluxes• Stratocumulus clouds in lowest 2 levels• Changes to autoconversion thresholds• Changes to relative humidity thresholds• Fall speed of droplets is function of effective radius

NETTOA = 4.5 Wm2

NETTOA = 0.53 Wm2

CAM 2CAM 3

CAM 3 – CAM 2



Global Aerosol Assimilation Climatology

Analysis NCAR CAMAerosol ForcingClimate Impacts

AVHRRAerosol Assimilation

Analysis

1995-2000



Addition of Prescribed Aerosol Forcing



Changes in Longwave Cooling Rates:New H2O Lines and Continuum

Change in LBL Cooling

Change in CAM Cooling



Global Decrease in Longwave Fluxes



Changes in Shortwave Heating Rates:New H2O Lines and Continuum

LBL Heating

CAM2 Heating

CAM2x Heating

Tropics



Global Increase in SW Heating Rates



Global Decrease in Surface Insolation



Higher Tropopause Temperatures

CAM 2CAM 3

CAM 3 – CAM 2



Lower Winter Land Surface Temperatures

CAM 2CAM 3

CAM 3 – CAM 2



Cloud Response to Tropical SST Variations

CAM 3



Cloud Radiative Forcing during ENSOs



Changes in Precipitation

CAM2/CLM2 CAM3/CLM2.2



Changes in Coupler & the Sea-Ice Model

• Coupler– Improved performance and scaling– Greater flexibility to add new fields– Significant DOE communication infrastructure– Faster multi-way communication to

components

• Sea Ice– Horizontal advection scheme– Addition of constant salinity in sea ice– Updates to albedos of snow and sea ice– Updates to ridging for thick ice categories



Global Energy BalanceControl Integration of CCSM3



Land Surface TemperaturesControl Integration of CCSM3



Ocean Surface TemperaturesControl Integration of CCSM3



Surface Wind StressControl Integration of CCSM3



Surface InsolationControl Integration of CCSM3



Sea-Ice ExtentControl Integration of CCSM3



Sea Ice Area and VolumeControl Integration of CCSM3



Climate Sensitivity of CCSM



Development Plans for CCSM, 2004-08



Development Plans for CCSM, 2004-08http://www.ccsm.ucar.edu/management

• CCSM Science Plan– Brief history of the CCSM program– Scientific accomplishments of CCSM2– Status of the CCSM– Planned climate experiments– Near-term program for understanding coupled model– Scientific questions needing major new capabilities

• CCSM Business Plan– Current resources– Vision for the future and resource requirements

http://www.ccsm.ucar.edu/management





CCSM: The Next Two Years

• Roadmap to the future– Climate sensitivity from IPCC studies– Process studies from GFDL collaboration, CPTs– Studies of higher resolution and

“benchmark” calculations– News physics/dynamics from Science Plan

• Integration of climate and chemistry– Ocean and land biogeochemistry– Prognostic aerosols– Tropospheric chemistry– WACCM– Tracers



The Evolution of CCSM



New Physics for High Resolution:Reduced Lateral Viscosity



Computational Demands for CCSM Science



Challenges

• Atmosphere– Convection– Cloud/radiation interactions– Incorporation of atmospheric chemistry

• Ocean– Eddy-resolving resolution– Mesoscale modeling of coastal zones

• Land– Urban land use change– Prognostic biogeography

• Sea Ice– Surface complexity on scales << grid spacing

the community climate system model

Documents