Main menu

Navigation

Removing atmospheric boundary conditions for aquaplanet simulations

3 posts / 0 new
Last post
herold.n@...
Removing atmospheric boundary conditions for aquaplanet simulations

Hi all, I want to run aquaplanet simulations but with my own SST file, using CAM4 in CESM1.2.2, thus the aquaplanet compsets won't do for me. From this site (http://hannahlab.org/cesm-aqua-planet-mode-with-prescribed-sst/) I was able to set up a run with a global fixed SST ocean. However, the anomalies between my sensitivity tests (with meridionally symmetric forcings) did not look very symmetric, leading me to realise I still have aerosol and other atmospheric forcing files that are prescribed at modern or pre-industrial values. My question is, how does one remove or zero out all of these boundary conditions?

Numerous posts I've found touch on this topic but are not very conclusive. For example, this post suggests a method to remove aerosol forcing (https://bb.cgd.ucar.edu/node/1001892) but I believe with a previous version of CAM4 as the parameters are slightly different. I would immensely appreciate a list (even partial) of the namelists/datafiles that need changing so that my aquaplanets have truly symmetric forcing.

olson

  In our experiments (with cam5 physics) we made copies of all the aerosol files
 and zeroed out all the constituents and then pointed to the new files in the
 namelist.  For example:

 ext_frc_specifier    = 'SO2 -> /glade/p/work/brianpm/aqua_aero/mam3_so2_elev_AQUA.nc',
         'bc_a1  -> /glade/p/work/brianpm/aqua_aero/mam3_bc_elev_AQUA.nc',
         'num_a1 -> /glade/p/work/brianpm/aqua_aero/mam3_num_a1_elev_AQUA.nc',
         'num_a2 -> /glade/p/work/brianpm/aqua_aero/mam3_num_a2_elev_AQUA.nc',
         'pom_a1 -> /glade/p/work/brianpm/aqua_aero/mam3_oc_elev_AQUA.nc',
         'so4_a1 -> /glade/p/work/brianpm/aqua_aero/mam3_so4_a1_elev_AQUA.nc',
         'so4_a2 -> /glade/p/work/brianpm/aqua_aero/mam3_so4_a2_elev_AQUA.nc'
 srf_emis_specifier   = 'DMS -> /glade/p/work/brianpm/aqua_aero/zero_DMS.nc',
         'SO2    -> /glade/p/work/brianpm/aqua_aero/mam3_so2_surf_AQUA.nc',
         'SOAG   -> /glade/p/work/brianpm/aqua_aero/mam3_soag_1.5_surf_AQUA.nc',
         'bc_a1  -> /glade/p/work/brianpm/aqua_aero/mam3_bc_surf_AQUA.nc',
         'num_a1 -> /glade/p/work/brianpm/aqua_aero/mam3_num_a1_surf_AQUA.nc',
         'num_a2 -> /glade/p/work/brianpm/aqua_aero/mam3_num_a2_surf_AQUA.nc',
         'pom_a1 -> /glade/p/work/brianpm/aqua_aero/mam3_oc_surf_AQUA.nc',
         'so4_a1 -> /glade/p/work/brianpm/aqua_aero/mam3_so4_a1_surf_AQUA.nc',
         'so4_a2 -> /glade/p/work/brianpm/aqua_aero/mam3_so4_a2_surf_AQUA.nc'


  All the above files contain 0's in the constituent fields

Jerry Olson

herold.n@...

Thanks for this Jerry. I've decided to stick with CAM5 since I was having issues with CAM4 running the 2000_CAM4_DLND%NULL_DICE%NULL_DOCN%DOM_SROF_SGLC_SWAV compset. I thought I'd document my experience here of running an aqua planet with my own custom SST profile.

Again, I started with CESM1.2.2 and instructions from http://hannahlab.org/cesm-aqua-planet-mode-with-prescribed-sst/. In addition to these I removed any seasonality in insolation by setting the following in user_nl_cpl.

&seq_infodata_inparm

        orb_mode = 'fixed_parameters'

        orb_obliq = 0.

        orb_eccen = 0.

        orb_mvelp = 0.

/

I also specified the aquaplanet ozone dataset https://www.earthsystemgrid.org/dataset/ucar.cgd.ccsm4.apeozone_cam3_5_5... so that ozone forcing is symmetric. This was done with the following namelist changes in user_nl_cam

&prescribed_ozone_nl

        prescribed_ozone_datapath              = '/short/w97/nkh562/CESM/user_inputdata/aquaplanet_CESM1.2.2'

        prescribed_ozone_file          = 'apeozone_cam3_5_54.nc'

        prescribed_ozone_name          = 'OZONE'

        prescribed_ozone_type = 'CYCLICAL'

        prescribed_ozone_cycle_yr = 1990

/

  This still leaves the effects of aerosol forcing which might explain the somewhat asymmetrical sturctures I see in the 200 mb geopotential height, in the monthly mean. But when averaging over several years this response becomes rather symmetric as expected. Following Jerry's comment above would remove this effect. I'm not sure if there are any other atmospheric boundary conditions I've missed, but for my purposes the behaviour of the model is good enough. I should also note that the aqua planet compsets (detailed here http://www.cesm.ucar.edu/models/simpler-models/aquaplanet.html) do not remove atmospheric boundary conditions like ozone etc. So you shouldn't expect a very symmetric climate from those compsets either.
Log in or register to post comments

Who's new

  • liuyiling.elaine@...
  • xuzq@...
  • alexander.koch....
  • abalos
  • bleisch@...