Gray Radiation Aquaplanet
Note, these instructions are valid for the 2023 CESM release cesm2_3_alpha16a and subsequent releases. For questions about this configuration, please post to the CESM simpler models forum.
This page describes how to run the gray radiation aquaplanet configuration within CESM. In this configuration, the physics package of CAM is replaced by the idealized gray radiation configuration following the protocol of Frierson et al (2006). The key features of this configuration are as follows:
- A simple gray radiation scheme is used. Solar radiation is specified as a flux at the surface as a function of latitude, so no solar radiation is absorbed by the atmosphere. There is a single longwave radiation band and the optical depth of the atmosphere for this band is specified as a function of latitude and pressure with a structure that aims to represent the structure of watervapor in the atmosphere.
- Water vapor is a prognostic variable in that it is sourced by evaporation from the surface, advected by the atmospheric dynamics and precipitates out when the atmosphere becomes saturated, but water vapor is not seen by the radiation scheme.
- When an atmospheric grid-box reaches saturation, large scale condensation occurs with condensed water vapor immediately rained out. This precipitation can be re-evaporated in the levels below, and the assumption is made that levels below must be saturated by re-evaporation for the precipitation to fall below that level.
- A simplified Monin-Obukhov similiarity theory and bulk formulae are used to determine the fluxes of heat, momentum and water vapor from the surface with the same drag coefficient used for each quantity (Equations 9, 10, and 11 of Frierson et al (2006)).
- The surface is a slab ocean with specified heat capacity and no representation of horizontal heat transport.
- There is no representation of clouds and a deep convection scheme is not currently implemented.
The gray radiation aquaplanet can be run using the compset FGRAYRAD. The following describes how to set up and run the gray radiation aquaplanet using a CESM release that is located in the directory $CESM.
Step 1: Create the gray radiation case
A gray radiation case with the finite volume dynamical core at 1 degree resolution can be set up by executing the following comment from the $CESM/cime/scripts directory
./create_newcase --case $CASEDIR --compset FGRAYRAD --res f09_f09_mg17 --run-unsupported
where $CASEDIR refers to the directory in which you'd like to make your case directory.
Note that the --run-unsupported flag will be necessary to use other grids with the FGRAYRAD compset.
Step 2: Set up the case
Set up the case by invoking the following command from within $CASEDIR
./case.setup
Step 3: Configure the case
Now you can configure your case by changing default settings and namelist variables. For example, the default simulation length is 5 days. This can be changed by altering the xml variables STOP_OPTION and STOP_N. For example, to change the run length to 6 months, the following command can be executed from within $CASEDIR ./xmlchange STOP_OPTION=nmonths, STOP_N=6 The output variables and frequencies can be modified by editing user_nl_cam and changing the fincl, mfilt, nhtfrq namelist parameters in the same manner as done for CAM. See instructions for modifying output fields for CAM.
The default history output is in the form of 30 day averages. To find out the history fields available in the gray radiation aquaplanet, you can preview the namelists from within $CASEDIR
./preview_namelists
Then look at ./CaseDocs/user_nl_cam and see the default settings for the fincl1 namelist variables. Alternatively, you could run a short simulation and examine the default output.
The default namelist parameters that control the gray radiation aquaplanet case can be altered by setting them in user_nl_cam. See the namelist parameters section below for further information on FGRAYRAD namelist parameters.
Step 4: Build the case
Build the case by executing the following command in $CASEDIR ./case.build If you're working on the NCAR supercomputer, you may also need to use the "qcmd" command qcmd -- ./case.build
Step 5: Submit the case
Submit the case by executing the following command in $CASEDIR ./case.submit
Namelist parameters in FGRAYRAD
The behavior of the FGRAYRAD compset is controlled by a variety of namelist parameters, most of which resemble those listed in Table 1 of Frierson et al (2006).
Namelist parameters controlling the behavior of the slab ocean:
- frierson_C0 is the ocean mixed-layer heat capacity. Default 1e7 J/K/m2.
- frierson_Tmin ($T_{min}$) controls the initialization of the ocean temperatures. It is the minimum temperature of the initialization profile. Default=271.
- frierson_Tdlt ($T_{dlt}$) controls the equator to pole SST difference of the initial ocean temperatures. Default=39.
- frierson_Twidth ($T_{w}$) controls the profile of the dependence of SST initial temperatures on latitude. Default=26$^{\circ}$.
The ocean temperatures are initialized with values equal to $T_{min} + T_{dlt}exp^{-0.5(\phi/T_{w})^{2}}$ where $\phi$ is latitude in degrees.
Namelist parameters controlling radiation:
- frierson_Albedo can be used to alter the Albedo. The net incoming solar flux at the surface is given by 1360*(1-frierson_Albedo). Default = 0.31.
- frierson_DeltaS ($\Delta_{s}$) controls the latitudinal variation of shortwave radiation. Default = 1.4
- frierson_Tau_eqtr ($\tau_{0e}$) controls the longwave optical depth at the equator. Default = 6
- frierson_Tau_pole ($\tau_{0p}$) controls the longwave optical depth at the pole. Default = 0.1
- frierson_LinFrac ($f_{l}$) controls the pressure dependence of the optical depth. Default = 0.1
The latitudinal structure of the optical depth is given by $\tau_{0}=\tau_{0e} + (\tau_{0p}-\tau_{0e})sin^{2}(\phi)$ where $\phi$ here is latitude in radians.
The overall structure of the optical depth is then given by $\tau = \tau_{0}\left[ f_{l}\left( \frac{p}{p_{s}} \right) + (1 - f_{l})\left( \frac{p}{p_{s}} \right)^{4} \right]$ where $p$ is the pressure and $p_{s}$ is surface pressure.
Parameters controlling the representation of surface turbulent fluxes:
- frierson_Z0 controls the roughness length for the surface drag calculation. Default = 3.21e-5m.
- frierson_Ri_c controls the critical Richardson number for stable mixing cutoff. Default = 1.
- frierson_Fb controls the fraction of the boundary layer that is considered to be a constant flux surface layer. Default = 0.1.
- frierson_Wind_min is a parameter that sets the minimum wind threshold for performing the surface flux calculations. It's added to avoid potential cases that could involve division by zero. Default = 1e-5.
Parameter controlling atmospheric water vapor:
- frierson_WetDryCoef is the parameter ($\xi$) used in Frierson et al (2006) to control the atmospheric water vapor content. It is a factor that is used to determine the saturation vapor pressure. A value of 0 represents the dry limit, a value of 10 represents the wet limit used in Frierson et al (2006).
Source code modification
For users who may want to change the source code of the gray radiation aquaplanet, the source code can be found at \$CESM/components/cam/src/physics/simple/frierson.F90 and frierson_cam.F90. These files can be copied into \$CASEDIR/SourceMods/src.cam and modified there and then these will be the version of the source code that will be compiled when running case.build.