SOCA background/model covariance. More...

Collaboration diagram for soca_covariance_mod::soca_cov:

[legend]

Public Member Functions
procedure	setup => soca_cov_setup
	Setup for the SOCA model's 3d error covariance matrices (B and Q_i) More...

procedure	delete => soca_cov_delete
	Delete for the SOCA model's 3d error covariance matrices. More...

procedure	mult => soca_cov_c_mult
	Apply convolution to an increment. More...

procedure	sqrt_c_mult => soca_cov_sqrt_c_mult
	Apply the square root of C to an increment. More...

procedure	getconv => soca_cov_get_conv
	Get the convolution operator needed for a specific field. More...

Public Attributes
type(bump_type), dimension(:), pointer	conv
	convolution op from bump More...

type(soca_state), pointer	bkg
	Background field (or first guess) More...

type(oops_variables)	vars
	Apply B to vars. More...

real(kind=kind_real), dimension(:), allocatable	pert_scale
	index matches "vars" More...

type(oops_variables), dimension(:), allocatable	conv_vars
	index mathces "conv" More...

Related Functions
(Note that these are not member functions.)
subroutine	soca_cov_setup (self, f_conf, geom, bkg, vars)
	Setup for the SOCA model's 3d error covariance matrices (B and Q_i) More...

subroutine	soca_cov_delete (self)
	Delete for the SOCA model's 3d error covariance matrices. More...

subroutine	soca_cov_get_conv (self, field, conv)
	Get the convolution operator needed for a specific field. More...

subroutine	soca_cov_c_mult (self, dx)
	Apply convolution to an increment. More...

subroutine	soca_cov_sqrt_c_mult (self, dx)
	Apply the square root of C to an increment. More...

subroutine	soca_bump_correlation (self, horiz_convol, geom, f_conf_bump, f_conf_domain, domain)
	Setup bump for horizontal convolution, Using rossby radiusbased correlation lengths. More...

subroutine	soca_2d_convol (dx, horiz_convol, geom)
	Apply bump 2D convolution. More...

subroutine	soca_2d_sqrt_convol (dx, horiz_convol, geom, pert_scale)
	Apply bump square root of C. More...

Detailed Description

SOCA background/model covariance.

Definition at line 29 of file soca_covariance_mod.F90.

Member Function/Subroutine Documentation

◆ delete()

procedure soca_covariance_mod::soca_cov::delete

Delete for the SOCA model's 3d error covariance matrices.

See also: soca_cov_delete

Definition at line 42 of file soca_covariance_mod.F90.

◆ getconv()

procedure soca_covariance_mod::soca_cov::getconv

Get the convolution operator needed for a specific field.

See also: soca_cov_get_conv

Definition at line 51 of file soca_covariance_mod.F90.

◆ mult()

procedure soca_covariance_mod::soca_cov::mult

Apply convolution to an increment.

See also: soca_cov_c_mult

Definition at line 45 of file soca_covariance_mod.F90.

◆ setup()

procedure soca_covariance_mod::soca_cov::setup

Setup for the SOCA model's 3d error covariance matrices (B and Q_i)

See also: soca_cov_setup

Definition at line 39 of file soca_covariance_mod.F90.

◆ sqrt_c_mult()

procedure soca_covariance_mod::soca_cov::sqrt_c_mult

Apply the square root of C to an increment.

See also: soca_cov_sqrt_c_mult

Definition at line 48 of file soca_covariance_mod.F90.

Friends And Related Function Documentation

◆ soca_2d_convol()

subroutine soca_2d_convol	(	real(kind=kind_real), dimension(:,:), intent(inout)	dx,
		type(bump_type), intent(inout)	horiz_convol,
		type(soca_geom), intent(in)	geom
	)

Used by soca_cov::mult()

Definition at line 365 of file soca_covariance_mod.F90.

◆ soca_2d_sqrt_convol()

subroutine soca_2d_sqrt_convol	(	real(kind=kind_real), dimension(:,:), intent(inout)	dx,
		type(bump_type), intent(inout)	horiz_convol,
		type(soca_geom), intent(in)	geom,
		real(kind=kind_real), intent(in)	pert_scale
	)

used by soca_cov::sqrt_C_mult()

Definition at line 392 of file soca_covariance_mod.F90.

◆ soca_bump_correlation()

subroutine soca_bump_correlation	(	class(soca_cov), intent(inout)	self,
		type(bump_type), intent(inout)	horiz_convol,
		type(soca_geom), intent(in)	geom,
		type(fckit_configuration), intent(in)	f_conf_bump,
		type(fckit_configuration), intent(in)	f_conf_domain,
		character(len=3), intent(in)	domain
	)

Used by soca_cov::setup()

Correlation lengths are calculated as follows : 1) rh = "base value" + rossby_radius * "rossby mult" 2) minimum value of "min grid mult" * grid_size is imposed 3) min/max are imposed based on "min value" and "max value" 4) converted from a gaussian sigma to Gaspari-Cohn cutoff distance

Parameters

[in,out] self The covariance structure

Definition at line 251 of file soca_covariance_mod.F90.

◆ soca_cov_c_mult()

subroutine soca_cov_c_mult	(	class(soca_cov), intent(inout)	self,
		type(soca_increment), intent(inout)	dx
	)

Parameters

[in,out]	self	The covariance structure
[in,out]	dx	Input: Increment, Output: C dx

Definition at line 172 of file soca_covariance_mod.F90.

◆ soca_cov_delete()

subroutine soca_cov_delete ( class(soca_cov), intent(inout) self )

Parameters

[in,out] self The covariance structure

Definition at line 124 of file soca_covariance_mod.F90.

◆ soca_cov_get_conv()

subroutine soca_cov_get_conv	(	class(soca_cov), intent(inout)	self,
		type(soca_field), intent(in), pointer	field,
		type(bump_type), intent(out), pointer	conv
	)

Exceptions

abor1_ftn aborts if trying to use a field not on the tracer grid

Parameters

[in]	field	The field that will be convolved
[out]	conv	pointer to resulting convolution

Definition at line 140 of file soca_covariance_mod.F90.

◆ soca_cov_setup()

subroutine soca_cov_setup	(	class(soca_cov), intent(inout)	self,
		type(fckit_configuration), intent(in)	f_conf,
		type(soca_geom), intent(in)	geom,
		type(soca_state), intent(in), target	bkg,
		type(oops_variables), intent(in)	vars
	)

This routine queries the configuration for the parameters that define the covariance matrix, and stores the relevant values in the error covariance structure.

Parameters

[in,out]	self	The covariance structure
[in]	f_conf	The configuration
[in]	geom	Geometry
[in]	bkg	Background
[in]	vars	List of variables

Definition at line 67 of file soca_covariance_mod.F90.

◆ soca_cov_sqrt_c_mult()

subroutine soca_cov_sqrt_c_mult	(	class(soca_cov), intent(inout)	self,
		type(soca_increment), intent(inout)	dx
	)

Exceptions

abor1_ftn aborts if no pertubation scales are given

Parameters

[in,out]	self	The covariance structure
[in,out]	dx	Input: Increment, Output: C^1/2 dx

Definition at line 204 of file soca_covariance_mod.F90.

Member Data Documentation

◆ bkg

type(soca_state), pointer soca_covariance_mod::soca_cov::bkg

Background field (or first guess)

Definition at line 31 of file soca_covariance_mod.F90.

◆ conv

type(bump_type), dimension(:), pointer soca_covariance_mod::soca_cov::conv

convolution op from bump

Definition at line 30 of file soca_covariance_mod.F90.

◆ conv_vars

type(oops_variables), dimension(:), allocatable soca_covariance_mod::soca_cov::conv_vars

index mathces "conv"

Definition at line 35 of file soca_covariance_mod.F90.

◆ pert_scale

real(kind=kind_real), dimension(:), allocatable soca_covariance_mod::soca_cov::pert_scale

index matches "vars"

Definition at line 34 of file soca_covariance_mod.F90.

◆ vars

type(oops_variables) soca_covariance_mod::soca_cov::vars

Apply B to vars.

Definition at line 32 of file soca_covariance_mod.F90.

The documentation for this type was generated from the following file:

src/soca/Covariance/soca_covariance_mod.F90

Public Member Functions

Public Attributes

Related Functions

Detailed Description

Member Function/Subroutine Documentation

◆ delete()

◆ getconv()

◆ mult()

◆ setup()

◆ sqrt_c_mult()

Friends And Related Function Documentation

◆ soca_2d_convol()

◆ soca_2d_sqrt_convol()

◆ soca_bump_correlation()

◆ soca_cov_c_mult()

◆ soca_cov_delete()

◆ soca_cov_get_conv()

◆ soca_cov_setup()

◆ soca_cov_sqrt_c_mult()

Member Data Documentation

◆ bkg

◆ conv

◆ conv_vars

◆ pert_scale

◆ vars