PUBLIC INTERFACE ~ PUBLIC ROUTINES ~ NAMELIST

Module ocean_barotropic_mod

Contact:  S.M. Griffies ,  Martin Schmidt (OBC) ,  Zhi Liang (OBC and halos) ,  Harper Simmons (tides)
Reviewers:  M.J. Harrison
Change History: WebCVS Log

OVERVIEW

Update the vertically integrated dynamics using a split-explicit algorithm.

This module time steps the vertically integrated dynamics using a predictor-Corrector with adjustable dissipation. This code uses a split-explicit method. There have been no no rigid lid available in MOM since MOM3. Treatment is provided for both depth-based Boussinesq and pressure-based non-Boussinesq vertical coordinates. Treatment is provided for both a Bgrid or Cgrid horizontal layout.

OTHER MODULES USED

           constants_mod
        diag_manager_mod
                 fms_mod
              fms_io_mod
         mpp_domains_mod
                 mpp_mod
        time_manager_mod
time_interp_external_mod
  ocean_bih_friction_mod
       ocean_domains_mod
  ocean_lap_friction_mod
  ocean_obc_barotrop_mod
           ocean_obc_mod
     ocean_operators_mod
    ocean_parameters_mod
         ocean_types_mod
          ocean_util_mod
     ocean_workspace_mod

PUBLIC INTERFACE

ocean_barotropic_init:
barotropic_diag_init:
eta_terms_diagnose_init:
eta_and_pbot_update:
eta_and_pbot_diagnose:
eta_and_pbot_tendency:
update_ocean_barotropic:
ocean_barotropic_forcing:
ocean_mass_forcing:
pred_corr_tropic_depth_bgrid:
pred_corr_tropic_depth_cgrid:
pred_corr_tropic_press_bgrid:
pred_corr_tropic_press_cgrid:
eta_smooth_diagnosed:
ocean_eta_smooth:
ocean_pbot_smooth:
barotropic_integrals:
barotropic_energy:
read_barotropic:
ocean_barotropic_restart:
ocean_barotropic_end:
maximum_convrhoud:
barotropic_chksum:
psi_compute:
eta_terms_diagnose:
eta_truncate:
eta_check:
tidal_forcing_init:
geoid_forcing_init:
get_tidal_forcing:
ideal_initialize_eta:
REMAP_BT_TO_BU_LOCAL:

PUBLIC ROUTINES

  1. ocean_barotropic_init

    DESCRIPTION
    Initialize the barotropic module


  2. barotropic_diag_init

    DESCRIPTION
    Initialize diagnostic indices for the module. Diagnose the static arrays.


  3. eta_terms_diagnose_init

    DESCRIPTION
    Initialize diagnostic indices for the eta_terms related diagnostics.


  4. eta_and_pbot_update

    DESCRIPTION
    Time step the surface height or pbot using a "big" time step. These fields are coincident in time with ocean tracers. NOTE: For pbot_t updates, we time step anompb for accuracy, then add rho0grav*ht to get pbot_t.


  5. eta_and_pbot_diagnose

    DESCRIPTION
    Diagnose surface height or pbot depending on the vertical coordinate. Note that dzt has been updated to taup1 before this routine is called, since we have already called update_tcell_thickness. Also, compute geodepth_zt in this routine. It is necessary to do do this step here, since for pressure coordinate models we do not know eta_t(taup1) until this routine. June 2011: note to smg: Consider changing taup1 to tau in rho, since we do not yet know the taup1 value. Not a fundamental issue, but may be cleaner. It will change bits, however.


  6. eta_and_pbot_tendency

    DESCRIPTION
    Compute tendency for surface height or bottom pressure. If use_blobs=.true., then we include the divergence from the Lagrangian system.


  7. update_ocean_barotropic

    DESCRIPTION
    Time step the external mode fields using a predictor-corrector scheme. Time average these fields to update the vertically integrated density weighted velocity (urhod,vrhod) and the time averaged surface height eta_t_bar or bottom pressure anompb_bar. NOTE: surface pressure gradient and gradient of anomalous bottom pressure are needed for the energy analysis. Also, if splitting=false or update_velocity_via_uprime=.false., use these for velocity update in ocean_velocity_mod. Include the tidal forcing if present. Include geoid perturbation if present. Use time averaged eta and pbot to ensure the stable pressure gradient for use with splitting=false or update_velocity_via_uprime=.false.


  8. ocean_barotropic_forcing

    DESCRIPTION
    Construct the vertically integrated forcing. This forcing is to be held constant over the barotropic timesteps. At the time of calling this routine, accel has the contributions from explicit-time forcing, except for the following: 1. Coriolis force is updated on the barotropic time steps when integrating the barotropic dynamics. So it should not be included in forcing_bt. 2. Contributions from smf and bmf are added to forcing_bt to allow for simpler handling of vertical friction implicitly in time. 3. The accel array is already thickness and density weighted, so a vertical density weighted integral is here just a vertical sum.


  9. ocean_mass_forcing

    DESCRIPTION
    Construct the vertically integrated mass source for evolution of the free surface and/or bottom pressure. Also construct the time tendency for the atmospheric pressure.


  10. pred_corr_tropic_depth_bgrid

    DESCRIPTION
    Integrate barotropic dynamics for "nts" timesteps using predictor-corrector. Assume depth-like vertical coordinate so solve for surface height. This scheme is more stable than leap_frog since it can run with longer time steps to resolve external mode gravity waves. It also provides some extra smoothing when pred_corr_gamma > 0 and so the options smooth_eta_t_bt_laplacian and smooth_eta_t_bt_biharmonic may not be needed. Time steps Bgrid layout of discrete fields.


  11. pred_corr_tropic_depth_cgrid

    DESCRIPTION
    Integrate barotropic dynamics for "nts" timesteps using predictor-corrector. Assume depth-like vertical coordinate so solve for surface height. Cgrid horizontal layout. This scheme is more stable than leap_frog since it can run with longer time steps to resolve external mode gravity waves. It also provides some extra smoothing when pred_corr_gamma > 0. We assume pred_corr_gamma > 0 in this routine.


  12. pred_corr_tropic_press_bgrid

    DESCRIPTION
    Integrate barotropic dynamics for "nts" timesteps using predictor-corrector. Assume pressure-like vertical coordinate so solve here for the bottom pressure. This scheme provides some smoothing of small spatial scales when pred_corr_gamma > 0. NOTE: the pressure gradient force is based on gradients of (pbot_t_bt - rho0*grav*ht) rather than gradients of pbot_t_bt. This approach aims to improve accuracy of the pressure force. Time steps Bgrid layout of discrete fields.


  13. pred_corr_tropic_press_cgrid

    DESCRIPTION
    Integrate barotropic dynamics for "nts" timesteps using predictor-corrector. Assume pressure-like vertical coordinate so solve here for the bottom pressure. This scheme provides some smoothing of small spatial scales with requirement that pred_corr_gamma > 0. NOTE: the pressure gradient force is based on gradients of (pbot_t_bt - rho0*grav*ht) rather than gradients of pbot_t_bt. This approach aims to improve accuracy of the pressure force. Time steps Cgrid layout of discrete fields.


  14. eta_smooth_diagnosed

    DESCRIPTION
    Smooth eta_t for case when running with pressure models and wish to have a diagnostic eta field smoothed. This option is most useful for the Bgrid, which has a checkerboard null mode when discretizing gravity waves. The Cgrid has no gravity wave noise so may not need this smoother.


  15. ocean_eta_smooth

    DESCRIPTION
    Compute tendency for smoothing eta and tracer. Use either a laplacian or a biharmonic smoothing operator. This smoothing option is most useful for the Bgrid, which has a checkerboard null mode when discretizing gravity waves. The Cgrid has no gravity wave noise so may not need this smoother. Recommend against using the biharmonic, since it is not a positive definite operator and so can lead to extrema. Biharmonic is retained for legacy purposes.


  16. ocean_pbot_smooth

    DESCRIPTION
    Compute tendency for diffusion of (pbot_t-pbot0) in both pbot_t and tracer. Need to consider tracer in order to maintain compability between tracer and mass conservation equations. Use either a laplacian or a biharmonic smoother. This smoothing option is most useful for the Bgrid, which has a checkerboard null mode when discretizing gravity waves. The Cgrid has no gravity wave noise so may not need this smoother. Recommend against using the biharmonic, since it is NOT a positive definite operator and so can lead to extrema.


  17. barotropic_integrals

    DESCRIPTION
    Compute area averaged fresh water and surface height and ocean mass.


  18. barotropic_energy

    DESCRIPTION
    Compute energetics of vertically integrated flow.


  19. read_barotropic

    DESCRIPTION
    Read in external mode fields from restart file.


  20. ocean_barotropic_restart

    DESCRIPTION
    Write out restart files registered through register_restart_file Call to reset_field_pointer only needed for fields with a time index.


  21. ocean_barotropic_end

    DESCRIPTION
    Write out external mode fields to restart file.


  22. maximum_convrhoud

    DESCRIPTION
    Compute maximum convergence(rho_ud,rho_vd).


  23. barotropic_chksum

    DESCRIPTION
    Compute checksum for external mode fields.


  24. psi_compute

    DESCRIPTION
    Compute quasi-barotropic streamfunctions for diagnostic purposes. When no fresh water and steady state, these two streamfunctions will be equal, and they will be equal to the rigid lid barotropic streamfunction in the Boussinesq case. Original algorithm: Stephen.Griffies 13MAR2007: Change units to 10^9 kg/s, which is a "mass Sv" This is the natural unit of transport for a mass-based vertical coordinate model. Updated Dec2009 to be compatible with tx_trans and ty_trans calculation.


  25. eta_terms_diagnose

    DESCRIPTION
    Diagnose various contributions to the sea level. WARNING: The steric diagnostics from this subroutine are confusing when evaluated in a Boussinesq model. The reason is that volume conserving Boussinesq models have spurious mass sources, which corrupt the bottom pressure signal. One needs to apply corrections to make sense of the Boussinesq models for purposes of studying mass budgets, including the local contribution to steric effects. contributions from dynamics, mass, and steric: eta_nonbouss = (eta_dynamic + eta_water + eta_source) + eta_steric = eta_nonsteric + eta_steric For PRESSURE_BASED vertical coordinates, eta_smooth_tend has already been computed in subroutine eta_smooth_diagnosed. We do not add this contribution to eta_nonbouss, since this piece is not part of the tendencies affecting bottom pressure. It is only added for cosmetic reasons. It is for this reason that eta_smooth is NOT included in the restart file. For calculation of the steric contribution, a single time step scheme is assumed, which is the recommended time stepping in MOM. For DEPTH_BASED models, the smoothing of eta is included in Ext_mode%source, so eta_smooth_tend is zero for depth-based models. For PRESSURE_BASED vertical coordinates, eta_nonbouss as computed in this routine is very close to the prognostic eta_t. Differences arise from any possible smoothing applied to the diagnosed eta_t.


  26. eta_truncate

    DESCRIPTION
    Truncate eta_t to keep dzt(1) + eta_t >= frac_crit_cell_height*dzt(1) and eta_t <= eta_max May be needed when run GEOPOTENTIAL models.


  27. eta_check

    DESCRIPTION
    Perform diagnostic check on top cell thickness. Useful when when use GEOPOTENTIAL vertical coordinate.


  28. tidal_forcing_init

    DESCRIPTION
    Initialize fields needed for lunar and solar tidal forcing.


  29. geoid_forcing_init

    DESCRIPTION
    Initialize fields needed for modifying the geoid, relative to the standard geoid.


  30. get_tidal_forcing

    DESCRIPTION
    Compute equilibrium tidal forcing.


  31. ideal_initialize_eta

    DESCRIPTION
    Idealized initial condition for eta.


  32. REMAP_BT_TO_BU_LOCAL

    DESCRIPTION
    Local version of the operator, needed here for initialization when read in eta information from an initialization file. Since barotropic is initialized prior to operators, we need to have this operator here locally.


NAMELIST

&ocean_barotropic_nml

write_a_restart
Set true to write a restart. False setting only for rare cases where wish to benchmark model without measuring the cost of writing restarts and associated chksums. Default is write_a_restart=.true.
[logical]
zero_tendency
If true, will not integrate the barotropic fields.
[logical]
zero_forcing_bt
Will set to zero all of the terms forcing the barotropic velocity.
[logical]
zero_nonlinear_forcing_bt
Will set to zero the nonlinear forcing terms, leaving only the smf and bmf terms to force the barotropic velocity.
[logical]
zero_eta_ic
To initialize eta_t to zero.
[logical]
zero_eta_t
To maintain eta_t at zero, but to allow other fields to evolve. For debugging. Default zero_eta_t=.false.
[logical]
zero_eta_u
To maintain eta_u at zero, but to allow other fields to evolve. For debugging. Default zero_eta_u=.false.
[logical]
zero_eta_tendency
To maintain deta_dt at zero. For debugging. Default zero_eta_t=.false.
[logical]
ideal_initial_eta
To initialize eta_t to an ideal profile. This option overrides all other initialization that may have occurred. Default=.false.
[logical]
ideal_initial_eta_amplitude
Amplitude for initializing eta with an ideal profile. Default ideal_initial_eta_amplitude = 5.0
[real, units: metre]
ideal_initial_eta_xwidth
Width in x-direction for sine-wave profile. Default xwidth=100e3
[real, units: metre]
ideal_initial_eta_ywidth
Width in y-direction for sine-wave profile. Default ywidth=100e3
[real, units: metre]
barotropic_time_stepping_A
Use the general approach from MOM4.0, in which the eta_t and pbot_t fields are updated with a big time step. This is the recommended approach for most applications that do not employ and open boundary condition. Default barotropic_time_stepping_A=.false.
[logical]
barotropic_time_stepping_B
Use the alternative approach in which we assume the barotropic scheme is a predictor-corrector, which is now the default in MOM. We use this assumption so that the eta_t and pbot_t fields are updated with a time average. This approach is used for open boundary condition applications. Default barotropic_time_stepping_A=.false.
[logical]
pred_corr_gamma
Dimensionless dissipation parameter for the preditor-corrector scheme. Setting pred_corr_gamma=0.0 reduces the scheme to a forward-backward, but it has been found to be unstable. So pred_corr_gamma > 0.0 is recommended. Note that pred_corr_gamma > 0.25 may be over-dissipated and so may go unstable. Default pred_corr_gamma=0.2.
[real, units: dimensionless]
smooth_eta_t_bt_laplacian
For spatially smoothing the eta_t field at each barotropic time step using a Laplacian operator. This option may not be necessary when pred_corr_gamma > 0.0, since the predictor-corrector approach has dissipation from pred_corr_gamma > 0.0. Also, smoothing is not needed in general for Cgrid MOM, since the gravity wave null mode only appears for the Bgrid. This option is only applicable for DEPTH_BASED vertical coordinates. Default smooth_eta_t_bt_laplacian=.false.
[logical]
smooth_eta_t_bt_biharmonic
For spatially smoothing the eta_t field at each barotropic time step using a biharmonic operator. May not be necessary when pred_corr_gamma > 0.0, since predictor-corrector has dissipation from pred_corr_gamma > 0.0. Also, smoothing is not needed in general for Cgrid MOM, since the gravity wave null mode only appears for the Bgrid. Applicable just for DEPTH_BASED vertical coordinates. WARNING: this operator is NOT positive definite, and so can produce spurious extrema. It is not generally recommended just for this reason. Default smooth_eta_t_bt_laplacian=.false.
[logical]
smooth_eta_t_laplacian
For spatially smoothing the eta_t field on the big time step by using a laplacian operator. For compatibility and global conservation, must also introduce a mixing to the thickness weighted tracer concentration in the k=1 cell. Applicable just for DEPTH_BASED vertical coordinates. Also, smoothing is not needed in general for Cgrid MOM, since the gravity wave null mode only appears for the Bgrid. Default mooth_eta_t_laplacian=.true.
[logical]
smooth_eta_t_biharmonic
For spatially smoothing the eta_t field on the big time step by using a biharmonic operator. For compatibility and global conservation, must also introduce a mixing to the thickness weighted tracer concentration in the k=1 cell. Applicable just for DEPTH_BASED vertical coordinates. Also, smoothing is not needed in general for Cgrid MOM, since the gravity wave null mode only appears for the Bgrid. WARNING: This operator is NOT positive definite, and so can produce spurious extrema. It is not recommended just for this reason. Default smooth_eta_t_biharmonic=.false.
[logical]
eta_offset
Uniform offset for use in determining the filter acting on tracer when smoothing the surface height. Default eta_offset=1e-12.
[real, units: metre]
smooth_eta_diag_laplacian
For spatially smoothing the diagnosed eta_t field using a laplacian operator. This option is used for PRESSURE_BASED vertical coordinates, in which case the free surface is diagnosed rather than prognosed. Also, smoothing is not needed in general for Cgrid MOM, since the gravity wave null mode only appears for the Bgrid. Default smooth_eta_diag_laplacian=.true.
[logical]
smooth_eta_diag_biharmonic
For spatially smoothing the diagnosed eta_t field using a biharmonic operator. This option is used for PRESSURE_BASED vertical coordinates, in which case the free surface is diagnosed rather than prognosed. Also, smoothing is not needed in general for Cgrid MOM, since the gravity wave null mode only appears for the Bgrid. Default smooth_eta_diag_biharmonic=.false.
[logical]
vel_micom_lap_diag
Velocity scale that is used for computing the MICOM Laplacian mixing coefficient used in the Laplacian smoothing of diagnosed surface height. Default vel_micom_lap_diag=0.2.
[real, units: m/sec]
vel_micom_bih_diag
Velocity scale that is used for computing the MICOM biharmonic mixing coefficient used in the biharmonic smoothing of diagnosed surface height. Default vel_micom_bih_diag=0.1.
[real, units: m/sec]
smooth_anompb_bt_laplacian
For spatially smoothing anomalous pbot_t at each barotropic time step using a Laplacian operator. May not be necessary when pred_corr_gamma > 0.0, since predictor-corrector has dissipation from pred_corr_gamma > 0.0. This option is applicable just for PRESSURE_BASED vertical coordinates. Also, smoothing is not needed in general for Cgrid MOM, since the gravity wave null mode only appears for the Bgrid. Default smooth_anompb_bt_laplacian=.false.
[logical]
smooth_anompb_bt_biharmonic
For spatially smoothing the anomalous pbot_t field at each barotropic time step using a biharmonic operator. May not be necessary when when pred_corr_gamma > 0.0, since predictor-corrector has dissipation from pred_corr_gamma > 0.0. This option is applicable just for PRESSURE_BASED vertical coordinates. Also, smoothing is not needed in general for Cgrid MOM, since the gravity wave null mode only appears for the Bgrid. WARNING: This operator is NOT positive definite, and so can produce spurious extrema. It is not recommended just for this reason. Default smooth_anompb_bt_biharmonic=.false.
[logical]
smooth_pbot_t_laplacian
For spatially smoothing pbot_t-pbot0 on the big time step using a laplacian operator. For compatibility and global conservation, must also introduce a mixing to the thickness weighted tracer concentration in the k=kbot cell. Applicable just for PRESSURE_BASED vertical coordinates. Also, smoothing is not needed in general for Cgrid MOM, since the gravity wave null mode only appears for the Bgrid. Default smooth_pbot_t_laplacian=.true.
[logical]
smooth_pbot_t_biharmonic
For spatially smoothing pbot_t-pbot0 on the big time step by using a biharmonic operator. For compatibility and global conservation, must also introduce a mixing to the thickness weighted tracer concentration in the k=kbot cell. Applicable just for PRESSURE_BASED vertical coordinates. Also, smoothing is not needed in general for Cgrid MOM, since the gravity wave null mode only appears for the Bgrid. WARNING: This operator is NOT positive definite, and so can produce spurious extrema. It is not recommended just for this reason. Default smooth_pbot_t_biharmonic=.false.
[logical]
smooth_pbot_t_biharmonic_legacy
For using an older version of the smooth_pbot_t_biharmonic scheme. The smooth_pbot_t_biharmonic_legacy option has a minor bug, but it is maintained in order to allow for backward compatible legacy simulations. It is not recommended for new simulations. To use it requires also setting smooth_pbot_t_biharmonic=.true. Default smooth_pbot_t_biharmonic_legacy=.false.
[logical]
pbot_offset
Uniform offset for use in determining the filter acting on tracer when smoothing the bottom pressure anomaly. Default pbot_offset=1e-12.
[real, units: Pa]
vel_micom_lap
Velocity scale that is used for computing the MICOM Laplacian mixing coefficient used in the Laplacian smoothing of surface height or anomalous bottom pressure. Default vel_micom_lap=0.05.
[real, units: m/sec]
vel_micom_bih
Velocity scale that is used for computing the MICOM biharmonic mixing coefficient used in the biharmonic smoothing of surface height or anomalous bottom pressure. Default vel_micom_bih=0.01.
[real, units: m/sec]
udrho_bt_lap
The vertically integrated horizontal momentum can be noisy on the Bgrid. It is therefore sometimes useful to add a smoothing operator to the barotropic time stepping. Here, we apply the laplacian friction as coded in the friction module using the vertically averaged isotropic viscosity as well as a background, and we do so on each barotropic time step. It is an expensive option. It is an option rarely used GFDL. This options will soon be removed from MOM. Default udrho_bt_lap=.false.
[logical]
udrho_bt_bih
The vertically integrated horizontal momentum on the Bgrid can be noisy. It is therefore sometimes useful to add a smoothing operator. Here, we apply the biharmonic friction as coded in the friction module using the vertically averaged isotropic viscosity as well as a background. Do so on each barotropic time step, which makes it an expensive option. This option is rarely used GFDL. Default udrho_bt_bih=.false. This options will soon be removed from MOM.
[logical]
udrho_lap
The vertically integrated horizontal momentum on the Bgrid can be noisy. It is therefore sometimes useful to add a smoothing operator. Here, we apply the laplacian friction as coded in the friction module using the vertically averaged isotropic viscosity as well as a background. Do so just on the baroclinic time step, so the option is less expensive than udrho_bt_lap. This options will soon be removed from MOM. Default udrho_lap=.false.
[logical]
udrho_bih
The vertically integrated horizontal momentum on the Bgrid can be noisy. It is therefore sometimes useful to add a smoothing operator. Here, we apply the biharmonic friction as coded in the friction module using the vertically averaged isotropic viscosity as well as a background. Do so just on the baroclinic time step, so the option is less expensive than udrho_bt_lap. This options will soon be removed from MOM. Default udrho_bih=.false.
[logical]
udrho_lap_vel_micom
Velocity scale that is used for computing the MICOM Laplacian mixing coefficient used in the Laplacian smoothing of udrho. This options will soon be removed from MOM. Default udrho_lap_vel_micom=.05
[real, units: m/sec]
udrho_bih_vel_micom
Velocity scale that is used for computing the MICOM biharmonic mixing coefficient used in the biharmonic smoothing of udrho. This options will soon be removed from MOM. Default udrho_bih_vel_micom=.01
[real, units: m/sec]
tidal_forcing_m2
Forces from lunar M2 tidal constituent. Default tidal_forcing_m2=.false.
[logical]
tidal_forcing_8
Forces from 8 lunar and solar tidal constituents. Default tidal_forcing_8=.false.
[logical]
tidal_forcing_ideal
For ideal tidal forcing, which has a bump configuration. Default tidal_forcing_ideal=.false.
[logical]
alphat
Dimensionless self-attraction and loading term. Used only when tidal_forcing=.true. Default alphat=0.948.
[real, units: dimensionless]
geoid_forcing
For modifying the geoid, implemented as a time independent tidal forcing. Need to read in a file to obtain the offset geoid profile. Default geoid_forcing=.false.
[logical]
truncate_eta
To truncate the surface height so to ensure positive thickness within the top cell. This method will not conserve volume or tracer. It is coded for cases when conservation is not critical but wish to run GEOPOTENTIAL models w/ large free surface height deviations, such as when running with tides and very fine vertical resolution. The preferred approach is to use zstar or pstart vertical coordinates. Default truncate_eta = .false..
[logical]
verbose_truncate
For verbose printout on truncate_eta
[logical]
frac_crit_cell_height
When use GEOPOTENTIAL vertical coordinate, the top model tracer grid cell has thickness dzt(i,j,1) = dzt(1) + eta_t(i,j). 0 < frac_crit_cell_height <= 1 sets the fraction of dzt(1) that is allowed prior to bringing the model down due to overly small dzt(i,j,1). Default frac_crit_cell_height=0.20.
[real, units: dimensionless]
eta_max
The maximum positive eta_t allowed when truncate_eta is true. Default eta_max = 5.0.
[real, units: meter]
barotropic_halo
Set barotropic_halo > 1 to use wide halo in the barotropic time step to improve the performance. In barotropic time step, most time is spent on mpp_update_domains. Use wide halo to decrease the number of mpp_update_domain calls and hence improve the performance. The default value is barotropic_halo=1, which is the older approach (non-wide halo). Users are encouraged to experiment with larger halos, as the model speedup can be tremendous.
[INTEGER]
do_bitwise_exact_sum
Set true to do bitwise exact global sum. When it is false, the global sum will be non-bitwise_exact, but will significantly increase efficiency. Default do_bitwise_exact_sum=.false.
[logical]
debug_this_module
Print out lots of diagnostics of use for debugging. Default debug_this_module=.false.
[logical]
verbose_init
For brief or full printout on initialization Default verbose_init=.true.
[logical]
diag_step
Frequency for output of ascii barotropic diagnostics. Setting diag_step=1 will compute diagnostics each time step and print to stdout. This setting is useful when developing a model in order to examine various budgets and stability issues. But when running production, one should set diag_step to a mucch larger number in order to reduce i/o and model cost. Default diag_step=-1, which means will not compute any of the online diagnostics.
[integer]

REFERENCES

  1. S.M. Griffies, R.C. Pacanowski, R.M. Schmidt, and V. Balaji Tracer Conservation with an Explicit Free Surface Method for Z-coordinate Ocean Models Monthly Weather Review (2001) vol 129 pages 1081--1098
  2. S.M. Griffies Fundamentals of Ocean Climate Models Princeton University Press (2004)
  3. S.M. Griffies, M.J. Harrison, R.C. Pacanowski, and A. Rosati A Technical Guide to MOM4 (2004)
  4. S.M. Griffies: Elements of MOM (2012)

top