Idealized modelling within the CESM framework
Many contributors (in alphabetical order): Alper Altuntas, Scott Bachman, Jim Benedict, Patrick Callaghan, Cheryl Craig, Gokhan Danabasoglu, Brain Dobbins, Brian Eaton, Andrew Gettelman, Steve Goldhaber, Christiane Jablonowski, Erik Kluzek Marysa Lague, Jean-Francois Lamarque, Peter Lauritzen, Sam Levis, Brian Medeiros, Kevin Reed, Bill Sacks, Isla Simpson, John Truesdale, Marana Vertenstein, Colin Zarzycki
International MIPs
+
This material is based upon work supported by the National Center for Atmospheric Research, which is a major facility sponsored by the National Science Foundation under Cooperative Agreement No. 1852977.
CESM components
Atmosphere
(CAM)
Ocean
(POP)
Sea Ice
(CICE)
Land ICE
(CISM)
Land
(CLM)
CESM components
Atmosphere
(CAM)
Ocean
(POP)
Sea Ice
(CICE)
Land ICE
(CISM)
Land
(CLM)
Coupler
Atmosphere
(CAM)
Atmosphere
(CAM)
Dynamics
Atmosphere
(CAM)
Dynamics
Radiative Transfer
Convection Scheme
Moist Processes
Cloud Physics
Radiative Transfer
Stresses due to sub-grid orography
Gravity Wave Drag
Surface Fluxes
Physical Parameterizations
Atmosphere
(CAM)
Dynamics
Radiative Transfer
Convection Scheme
Moist Processes
Cloud Physics
Radiative Transfer
Stresses due to sub-grid orography
Gravity Wave Drag
Surface Fluxes
Physical Parameterizations
Land (CLM)
Prescribed SSTs
Prescribed Sea Ice
WHY?
Problems
Problems
CESM is complicated. Everything is changing all at once
Problems
CESM is complicated. Everything is changing all at once
The system is typically in a quasi-equilibrium/balanced state obeying the various balance’s it is supposed to (Energy, Momentum, Moisture)
Problems
CESM is complicated. Everything is changing all at once
The system is typically in a quasi-equilibrium/balanced state obeying the various balance’s it is supposed to (Energy, Momentum, Moisture)
All components are strongly coupled and interacting to ensure these balances are maintained. One thing changes, everything else responds, making it hard to establish causal relationships.
Problems
CESM is complicated. Everything is changing all at once
The system is typically in a quasi-equilibrium/balanced state obeying the various balance’s it is supposed to (Energy, Momentum, Moisture)
All components are strongly coupled and interacting to ensure these balances are maintained. One thing changes, everything else responds, making it hard to establish causal relationships.
To obtain the solution we had to use a large supercomputer 🡪 speaks to the complexity of the processes involved.
How can we pull it all apart and understand it?
How can we pull it all apart and understand it?
Detailed diagnosis of model output
How can we pull it all apart and understand it?
Detailed diagnosis of model output
Using simplified versions of CESM
How can we pull it all apart and understand it?
Detailed diagnosis of model output
Using simplified versions of CESM
Peforming idealized experiments with the comprehensive version of CESM
How can we pull it all apart and understand it?
Detailed diagnosis of model output
Using simplified versions of CESM
Peforming idealized experiments with the comprehensive version of CESM
How can we pull it all apart and understand it?
Detailed diagnosis of model output
Using simplified versions of CESM
Peforming idealized experiments with the comprehensive version of CESM
The capacity to run idealized models within CESM is growing
Simpler models website: https://www.cesm.ucar.edu/models/simple
Simpler models are…
…stripped down versions of CESM that only contain certain components and/or idealized representation of certain components.
Simpler models are…
…stripped down versions of CESM that only contain certain components and/or idealized representation of certain components.
PRO’s
CON’s
Simpler models are…
…stripped down versions of CESM that only contain certain components and/or idealized representation of certain components.
PRO’s
CON’s
Easy to perturb
Simpler models are…
…stripped down versions of CESM that only contain certain components and/or idealized representation of certain components.
PRO’s
CON’s
Easy to perturb
Allow for idealized experiments to identify causal pathways
Simpler models are…
…stripped down versions of CESM that only contain certain components and/or idealized representation of certain components.
PRO’s
CON’s
Easy to perturb
Allow for idealized experiments to identify causal pathways
Cheap
Simpler models are…
…stripped down versions of CESM that only contain certain components and/or idealized representation of certain components.
PRO’s
CON’s
Easy to perturb
Allow for idealized experiments to identify causal pathways
Cheap
Allows for parameter sweeps to identify sensitivities
Simpler models are…
…stripped down versions of CESM that only contain certain components and/or idealized representation of certain components.
PRO’s
CON’s
Easy to perturb
Allow for idealized experiments to identify causal pathways
Cheap
Allows for parameter sweeps to identify sensitivities
Less realistic
Simpler models are…
…stripped down versions of CESM that only contain certain components and/or idealized representation of certain components.
PRO’s
CON’s
Easy to perturb
Allow for idealized experiments to identify causal pathways
Cheap
Allows for parameter sweeps to identify sensitivities
Less realistic
Always keep your eye on the real world/full CESM
Atmospheric Simpler Models
The atmospheric model hierarchy
Dry Dynamical Core
Dynamical Core with Idealized moisture
Gray Radiation Aquaplanet
RCE World
Aquaplanet
CAM
Single Column Atmospheric Model (SCAM)
Shallow Water
Barotropic Models
Stationary Wave Models
Nudging
Cloud Locking
www.cesm.ucar.edu/models/simple/
The atmospheric model hierarchy
Dry Dynamical Core
Gray Radiation Aquaplanet
RCE World
Aquaplanet
CAM
Single Column Atmospheric Model (SCAM)
Shallow Water
Barotropic Models
Stationary Wave Models
Available CESM2.0 and later
Available CESM2.1 and later
Available CESM2.3_alpha16a and later
www.cesm.ucar.edu/models/simple/
Available CESM2.1.3 and later
Gray Radiation Aquaplanet
Dynamical Core with Idealized moisture
Nudging
Cloud Locking
Aspirational
The atmospheric model hierarchy
Dry Dynamical Core
Gray Radiation Aquaplanet
RCE World
Aquaplanet
CAM
Single Column Atmospheric Model (SCAM)
Shallow Water
Barotropic Models
Stationary Wave Models
Available CESM2.0 and later
Available CESM2.1 and later
Available CESM2.3_alpha16a and later
www.cesm.ucar.edu/models/simple/
Available CESM2.1.3 and later
Gray Radiation Aquaplanet
Dynamical Core with Idealized moisture
Nudging
Cloud Locking
Aspirational
Atmosphere
(CAM)
Dynamics
Radiative Transfer
Convection Scheme
Moist Processes
Cloud Physics
Radiative Transfer
Stresses due to sub-grid orography
Gravity Wave Drag
Surface Fluxes
Physical Parameterizations
Atmosphere
(CAM)
Dynamics
Radiative Transfer
Convection Scheme
Moist Processes
Cloud Physics
Radiative Transfer
Stresses due to sub-grid orography
Gravity Wave Drag
Surface Fluxes
Physical Parameterizations
Atmosphere
(CAM)
Dynamics
Newtonian relaxation of the temperature field toward a specified equilibrium profile
Linear drag on wind at the lowest levels
The atmospheric model hierarchy
Dry Dynamical Core
Gray Radiation Aquaplanet
RCE World
Aquaplanet
CAM
Single Column Atmospheric Model (SCAM)
Shallow Water
Barotropic Models
Stationary Wave Models
Available CESM2.0 and later
Available CESM2.1 and later
Available CESM2.3_alpha16a and later
www.cesm.ucar.edu/models/simple/
Available CESM2.1.3 and later
Gray Radiation Aquaplanet
Dynamical Core with Idealized moisture
Nudging
Cloud Locking
Aspirational
Dry Dynamical Core: https://www.cesm.ucar.edu/models/simple/held-suarez
All physical parameterizations replaced by Newtonian relaxation of the temperature field toward a zonally symmetric equilibrium temperature profile and linear drag on the near surface winds, following Held and Suarez (1994).
Currently runs with all dynamical cores (Eulerian, Finite Volume, Spectral Element, MPAS, FV3)
Good for dry dynamics. Can easily perturb the temperature
The atmospheric model hierarchy
Dry Dynamical Core
Gray Radiation Aquaplanet
RCE World
Aquaplanet
CAM
Single Column Atmospheric Model (SCAM)
Shallow Water
Barotropic Models
Stationary Wave Models
Available CESM2.0 and later
Available CESM2.1 and later
Available CESM2.3_alpha16a and later
www.cesm.ucar.edu/models/simple/
Available CESM2.1.3 and later
Gray Radiation Aquaplanet
Dynamical Core with Idealized moisture
Nudging
Cloud Locking
Aspirational
Dry Dynamical Core: https://www.cesm.ucar.edu/models/simple/held-suarez
All physical parameterizations replaced by Newtonian relaxation of the temperature field toward a zonally symmetric equilibrium temperature profile and linear drag on the near surface winds, following Held and Suarez (1994).
Currently runs with all dynamical cores (Eulerian, Finite Volume, Spectral Element, MPAS, FV3)
Good for dry dynamics. Can easily perturb the temperature
The atmospheric model hierarchy
Dry Dynamical Core
Gray Radiation Aquaplanet
RCE World
Aquaplanet
CAM
Single Column Atmospheric Model (SCAM)
Shallow Water
Barotropic Models
Stationary Wave Models
Available CESM2.0 and later
Available CESM2.1 and later
Available CESM2.3_alpha16a and later
www.cesm.ucar.edu/models/simple/
Available CESM2.1.3 and later
Gray Radiation Aquaplanet
Dynamical Core with Idealized moisture
Nudging
Cloud Locking
Aspirational
Atmosphere
(CAM)
Dynamics
Newtonian relaxation of the temperature field toward a specified equilibrium profile
Linear drag on wind at the lowest levels
Atmosphere
(CAM)
Dynamics
Newtonian relaxation of the temperature field toward a specified equilibrium profile
Linear drag on wind at the lowest levels
Water covered Earth
Prescribed SSTs
Evaporation
Heating associated with precipitation
The atmospheric model hierarchy
Dry Dynamical Core
Gray Radiation Aquaplanet
RCE World
Aquaplanet
CAM
Single Column Atmospheric Model (SCAM)
Shallow Water
Barotropic Models
Stationary Wave Models
Available CESM2.0 and later
Available CESM2.1 and later
Available CESM2.3_alpha16a and later
www.cesm.ucar.edu/models/simple/
Available CESM2.1.3 and later
Gray Radiation Aquaplanet
Dynamical Core with Idealized moisture
Nudging
Cloud Locking
Aspirational
Moist Held-Suarez (Thatcher and Jablonowski 2016): https://www.cesm.ucar.edu/models/simple/moist-held-suarez
Like the dry dynamical core but with a representation of the large scale condensation of moisture and associated diabatic heating.
Water covered Earth, prescribed SST profile. Representation of surface sensible and latent heat flux using bulk formulae.
Newtonian relaxation of the temperature field.
Moisture is advected by the large scale circulation, consensus when it reaches saturation and immediately precipitated with an associated diabatic heating.
Good for dynamical studies involving the interaction between moisture and the large scale flow.
Precipitation in moist Held-Suarez
The atmospheric model hierarchy
Dry Dynamical Core
Gray Radiation Aquaplanet
RCE World
Aquaplanet
CAM
Single Column Atmospheric Model (SCAM)
Shallow Water
Barotropic Models
Stationary Wave Models
Available CESM2.0 and later
Available CESM2.1 and later
Available CESM2.3_alpha16a and later
www.cesm.ucar.edu/models/simple/
Available CESM2.1.3 and later
Gray Radiation Aquaplanet
Dynamical Core with Idealized moisture
Nudging
Cloud Locking
Aspirational
Atmosphere
(CAM)
Dynamics
Newtonian relaxation of the temperature field toward a specified equilibrium profile
Linear drag on wind at the lowest levels
Water covered Earth
Prescribed SSTs
Evaporation
Heating associated with precipitation
Atmosphere
(CAM)
Dynamics
Bulk formulae for surface drag and sensible and latent heat fluxes
Water covered Earth
Prescribed SSTs
Evaporation
Heating associated with precipitation
Radiative Transfer
A simplified radiation scheme. Incoming shortwave. One longwave band with a specified longwave absorber. No clouds. Radiation scheme is not impacted by the moisture
The atmospheric model hierarchy
Dry Dynamical Core
Gray Radiation Aquaplanet
RCE World
Aquaplanet
CAM
Single Column Atmospheric Model (SCAM)
Shallow Water
Barotropic Models
Stationary Wave Models
Available CESM2.0 and later
Available CESM2.1 and later
Available CESM2.3_alpha16a and later
www.cesm.ucar.edu/models/simple/
Available CESM2.1.3 and later
Gray Radiation Aquaplanet
Dynamical Core with Idealized moisture
Nudging
Cloud Locking
Aspirational
Gray Radiation Aquaplanet (coming soon)
Good for idealized studies of the interactions between the circulation and radiation and moisture
The atmospheric model hierarchy
Dry Dynamical Core
Gray Radiation Aquaplanet
RCE World
Aquaplanet
CAM
Single Column Atmospheric Model (SCAM)
Shallow Water
Barotropic Models
Stationary Wave Models
Available CESM2.0 and later
Available CESM2.1 and later
Available CESM2.3_alpha16a and later
www.cesm.ucar.edu/models/simple/
Available CESM2.1.3 and later
Gray Radiation Aquaplanet
Dynamical Core with Idealized moisture
Nudging
Cloud Locking
Aspirational
The atmospheric model hierarchy
Dry Dynamical Core
Gray Radiation Aquaplanet
RCE World
Aquaplanet
CAM
Single Column Atmospheric Model (SCAM)
Shallow Water
Barotropic Models
Stationary Wave Models
Available CESM2.0 and later
Available CESM2.1 and later
Available CESM2.3_alpha16a and later
www.cesm.ucar.edu/models/simple/
Available CESM2.1.3 and later
Gray Radiation Aquaplanet
Dynamical Core with Idealized moisture
Nudging
Cloud Locking
Aspirational
Radiative Convective Equilibrium (RCE) world: https://www.cesm.ucar.edu/models/simple/rce
Compatible with the RCEMIP protocol.
No rotation, uniform and constant insolation
Uniform prescribed SSTs
Planetary rotation and solar zenith angle can be specified.
295K
305K
Reed et al (2021)
The atmospheric model hierarchy
Dry Dynamical Core
Gray Radiation Aquaplanet
RCE World
Aquaplanet
CAM
Single Column Atmospheric Model (SCAM)
Shallow Water
Barotropic Models
Stationary Wave Models
Available CESM2.0 and later
Available CESM2.1 and later
Available CESM2.3_alpha16a and later
www.cesm.ucar.edu/models/simple/
Available CESM2.1.3 and later
Gray Radiation Aquaplanet
Dynamical Core with Idealized moisture
Nudging
Cloud Locking
Aspirational
The atmospheric model hierarchy
Dry Dynamical Core
Gray Radiation Aquaplanet
RCE World
Aquaplanet
CAM
Single Column Atmospheric Model (SCAM)
Shallow Water
Barotropic Models
Stationary Wave Models
Available CESM2.0 and later
Available CESM2.1 and later
Available CESM2.3_alpha16a and later
www.cesm.ucar.edu/models/simple/
Available CESM2.1.3 and later
Gray Radiation Aquaplanet
Dynamical Core with Idealized moisture
Nudging
Cloud Locking
Aspirational
Aquaplanet: https://www.cesm.ucar.edu/models/simple/aquaplanet
Full CAM4, CAM5 or CAM6 physics.
Water covered Earth.
Prescribed SSTs or slab ocean.
Spectra of equatorial waves in the CAM5 aquaplanet (Medeiros et al 2016)
The atmospheric model hierarchy
Dry Dynamical Core
Gray Radiation Aquaplanet
RCE World
Aquaplanet
CAM
Single Column Atmospheric Model (SCAM)
Shallow Water
Barotropic Models
Stationary Wave Models
Available CESM2.0 and later
Available CESM2.1 and later
Available CESM2.3_alpha16a and later
www.cesm.ucar.edu/models/simple/
Available CESM2.1.3 and later
Gray Radiation Aquaplanet
Dynamical Core with Idealized moisture
Nudging
Cloud Locking
Aspirational
Single Column Atmospheric Model (SCAM), Gettelman et al 2019: https://www.cesm.ucar.edu/models/simple/scam
Full CAM physics.
Simulation of a single column. Large scale tendencies prescribed from either observations or a simulation.
RCE and Weak Temperature Gradient parameterizations of the large scale circulation are being implemented (U. Miami, Columbia)
Cloud fraction in SCAM6 and CAM6 (Gettelman et al 2019)
Useful for parameter sensitivity studies to explore how the physical parameterizations behave under different climates and different parameter settings.
Cheaper comprehensive atmospheric physics options
Maintaining CAM4 and CAM5 functionality within CESM
Click to add footer
NSF CSSI award 2311376
Isla Simpson Julio Bacmeister Peter Lauritzen Brian Medeiros Haipeng Lin. Adam Herrington
CAM4
CAM5
CAM6
CAM7
(2010)
(2012)
(2017)
Next Generation (~2026)
Physics packages within CESM:
Computational performance metrics based on a 1 month run a 1 degree horizontal resolution in an aquaplanet configuration using the FV dycore with 4 nodes (144 processors)
Land Simpler Models
SLIM (The Simple Land Interface Model)
Marysa Laguё
Abby Swann
Gordon Bonan
Erik Kluzek
CLM5
SLIM
www.marysalague.com
https://www.cesm.ucar.edu/models/cesm2/land/CLM50_Tech_Note.pdf
Solves linearized bulk surface energy budget coupled with soil temperatures and bucket hydrology.
Prescribed: Albedo’s, surface emissivity, soil conductivity and heat capacity, bucket capacity, evaporative resistance, vegetation height (aerodynamic roughness).
Allows for much more flexibility in prescribing land surface properties as opposed to letting them emerge as a result of the biophysics in CLM.
SLIM (The Simple Land Interface Model)
Marysa Laguё
Abby Swann
Gordon Bonan
Erik Kluzek
CLM5
SLIM
www.marysalague.com
https://www.cesm.ucar.edu/models/cesm2/land/CLM50_Tech_Note.pdf
Solves linearized bulk surface energy budget coupled with soil temperatures and bucket hydrology.
Prescribed: Albedo’s, surface emissivity, soil conductivity and heat capacity, bucket capacity, evaporative resistance, vegetation height (aerodynamic roughness).
Allows for much more flexibility in prescribing land surface properties as opposed to letting them emerge as a result of the biophysics in CLM.
Coming Soon
(It works, but hasn’t been released yet)
Ocean simpler models
The Pencil Model
Young-Oh Kwon
Gokhan Danabasoglu
Single column ocean model at each grid point.
No large scale ocean dynamics (prescribed tendencies of temperature and salinity to maintain climatology close to the coupled model).
Representation of mixed layer physics, prognostic mixed layer depth etc.
Pre-industrial simulations with this configuration are about to start
and others
Note: Pencil model development so far has been with the POP ocean model. CESM3 will use MOM ocean model. Additional resources will be needed to further the development of this in MOM, but we’d be glad to hear about interest in this.
Coupled Idealized Modelling
Coupled Idealized Modelling Tools – coming soon
NSF CSSI award 2004575
Scott Bachman Isla Simpson Gokhan Danabsoglu Mariana Vertenstein Alper Altuntas Brian Dobbins Sam Levis Bill Sacks
Aim: To allow users to easily set up their own idealized coupled configurations or atmosphere-land configurations
Ridge World
visualCaseGen
Alper Altuntas Sam Levis
A Jupyter-based GUI that will streamline the creation and configuration of CESM3 cases
(Currently works in CESM3 development tags)
esmci.github.io/visualCaseGen/
https://esmci.github.io/visualCaseGen/
�
International MIPs
Analysis from MIPs feeds into the Intergovernmental Panel on Climate Change reports
2021
2013
2007
2001
1995
1992
1990
CMIP3
CMIP5
CMIP6
Opportunity to:
(1) Intercompare models all on an equal footing – same experimental protocol.
(2) Update projections for the future of the climate system with the latest model generations to feed into assessment reports.
Coupled Model Intercomparison Project (CMIP)
Experiments can be broadly categorized into four categories
Coupled Model Intercomparison Project (CMIP)
Experiments can be broadly categorized into four categories
Characterizing model behavior
e.g., pre-industrial control, abrupt4xCO2, historical, AMIP
Coupled Model Intercomparison Project (CMIP)
Experiments can be broadly categorized into four categories
Characterizing model behavior
e.g., pre-industrial control, abrupt4xCO2, historical, AMIP
Interpreting the historical record
DAMIP single forcing experiments, RFMIP simulations, AerChemMIP LMIP
Coupled Model Intercomparison Project (CMIP)
Experiments can be broadly categorized into four categories
Characterizing model behavior
e.g., pre-industrial control, abrupt4xCO2, historical, AMIP
Interpreting the historical record
DAMIP single forcing experiments, RFMIP simulations, AerChemMIP LMIP
Predicting the future
ScenarioMIP, DCPP
Coupled Model Intercomparison Project (CMIP)
Experiments can be broadly categorized into four categories
Characterizing model behavior
e.g., pre-industrial control, abrupt4xCO2, historical, AMIP
Interpreting the historical record
DAMIP single forcing experiments, RFMIP simulations, AerChemMIP LMIP
Predicting the future
ScenarioMIP, DCPP
Targeted experiments for process understanding and/or to explore particular climate states
PMIP, GeoMIP, C4MIP, CFMIP experiments
Plans for CMIP7
The DECK
Entry card for CMIP: key experiments that are needed to characterize model behavior
Plans for CMIP7
The DECK
Entry card for CMIP: key experiments that are needed to characterize model behavior
Plans for CMIP7
The DECK
Entry card for CMIP: key experiments that are needed to characterize model behavior
The Assessment FastTrack
Priority experiments required for analysis that will feed into AR7 (timeline = Jan 2027)
Plans for CMIP7
The DECK
Entry card for CMIP: key experiments that are needed to characterize model behavior
The Assessment FastTrack
Priority experiments required for analysis that will feed into AR7 (timeline = Jan 2027)
Community MIPs
Other experiments proposed by the community that can occur on a more flexible timeline
Plans for CMIP7
The DECK
Entry card for CMIP: key experiments that are needed to characterize model behavior
The Assessment FastTrack
Priority experiments required for analysis that will feed into AR7 (timeline = Jan 2027)
Community MIPs
Other experiments proposed by the community that can occur on a more flexible timeline
Conclusions
Simpler models are valuable tools to gain a process level understanding of the behavior of the real world and/or comprehensive CESM and an understanding of sensitivities within the climate system.
Many of them are cheaper to run. Some of them you can even run on your own laptop.
They are also well documented with comprehensive instructions for how to modify them.
See the simpler models website: https://www.cesm.ucar.edu/models/simple
Post query’s to the bulletin board: https://bb.cgd.ucar.edu/cesm/forums/simpler-models.161/
My email address: islas@ucar.edu
International MIPs are a great opportunity to (a) test and intercompare models, (b) provide information that can feed into policymaking/adaptation measures, (c) interpret the historical record, (d) come up with fundamental understanding of processes in the climate system and their representation in models.
Extra Slides
VisualCaseGen Example
The starting point: GUI to choose your components
Click to add footer
Alper Altuntas
The GUI will allow you to choose your components and set up your component set
For idealized simulations with user defined geometries, the GUI will guide users through the different aspects that are needed for each component and to couple them together
The Pencil Model – coming soon
Young-Oh Kwon
Ivan Lima
Gokhan Danabasoglu
+ others…
Single column ocean model at each grid point.
No large scale ocean dynamics (prescribed tendencies of temperature and salinity to maintain climatology close to the coupled model).
Representation of mixed layer physics, prognostic mixed layer depth etc.
Methodology currently being refined and long simulations about to begin.