mom_stoch_eos module reference¶

Provides the ocean stochastic equation of state.

Data Types¶

mom_stoch_eos_cs

Describes parameters of the stochastic component of the EOS correction, described in Stanley et al.

Functions/Subroutines¶

`mom_stoch_eos_init()`	Initializes MOM_stoch_eos module, returning a logical indicating whether this module will be used.
`stoch_eos_register_restarts()`	Register fields related to the stoch_EOS module for resarts.
`mom_stoch_eos_run()`	Generates a pattern in space and time for the ocean stochastic equation of state.
`post_stoch_eos_diags()`	Write out any diagnostics related to this module.
`mom_calc_vart()`	Computes a parameterization of the SGS temperature variance.

Detailed Description¶

This module provides the foundation of the Stanley parameterization (stanley2020) for correcting the computation of density. Density is not a prognostic variable in MOM6; it is computed for various purposes in various places. The correction to this calculation provided by this module has been implemented in some places where density is used, but not all.

To use the correction, first set STOCH_EOS=True. Then, choose the constant c from (25) of stanley2020. This is controlled using STANLEY_COEFF. Setting a negative value will result in an error. stanley2020 found a value of 0.2 offline, coarsening from 0.1 to 1 degree resolution. kenigson2022 proposed a value of 0.5 in a 2/3 degree resolution model.

Whether the correction is deterministic or stochastic can be controlled using the variable STANLEY_A. Setting this to 0.0 uses the deterministic version, while a value of 1.0 produces the stochastic version. Reducing from 1 to 0 smoothly transitions from stochastic to deterministic.

To turn the correction on in various parts of the code, use * USE_STANLEY_PGF=True for the pressure gradient force (cf. kenigson2022)

USE_STANLEY_ISO=True to correct the computation of isopycnal slopes (used in many places)
USE_STANLEY_GM=True to use the parameterization within GM (cf. agarwal2023)
USE_STANLEY_ML=True to use the parameterization within the mixed-layer restratification parameterization. It applies to both the OM4 and Bodner schemes. (cf. agarwal2023)

For ensemble simulations, the random number generator seed can be controlled using the parameter SEED_STOCH_EOS

Type Documentation¶

type mom_stoch_eos/mom_stoch_eos_cs¶

Describes parameters of the stochastic component of the EOS correction, described in Stanley et al. JAMES 2020.

Type fields:

% id_stoch_eos [integer] :: Diagnostic IDs.
% id_stoch_phi [integer] :: Diagnostic IDs.
% id_tvar_sgs [integer] :: Diagnostic IDs.
% l2_inv [real(:,:),allocatable] :: One over sum of the T cell side side lengths squared [L-2 ~> m-2].
% rgauss [real(:,:),allocatable] :: nondimensional random Gaussian [nondim]
% tfac [real] :: Nondimensional decorrelation time factor, ~1/3.7 [nondim].
% amplitude [real] :: Nondimensional standard deviation of Gaussian [nondim].
% seed [integer] :: PRNG seed.
% rn_cs [type( prng )] :: PRNG control structure.
% pattern [real(:,:),allocatable] :: Random pattern for stochastic EOS [nondim].
% phi [real(:,:),allocatable] :: temporal correlation stochastic EOS [nondim]
% use_stoch_eos [logical] :: If true, use the stochastic equation of state (Stanley et al. 2020)
% stanley_coeff [real] :: Coefficient correlating the temperature gradient and SGS T variance [nondim]; if <0, turn off scheme in all codes.
% stanley_a [real] :: a in exp(aX) in stochastic coefficient [nondim]
% kappa_smooth [real] :: A diffusivity for smoothing T/S in vanished layers [H Z T-1 ~> m2 s-1 or kg m-1 s-1].