horiz_interp_type_mod

Data Types
interface	assignment(=)
	Interface to override the "=" operator to a horiz_interp_eq that will copy the Interp instance of horiz_interp_type with Interp2 = Interp1. More...
type	horiz_interp_type
	Datatype holding interpolation weights, mapping indices, and metadata for horizontal interpolation. All real members are stored in horizInterpReals8_type if the grid and data are represented in 64-bit floating point precision or horizInterpReals4_type if the grid and data are represented in 32-bit floating point precision. Only one type, horizInterpReals4_type or horizInterpReals8_type, is allocated and used for interpolation. More...
type	horizinterpreals4_type
	Nested derived type in horiz_interp_type that holds 32-bit interpolation weights, and metadata. If 32-bit grid and data arrays are provided to horiz_interp, InterphorizInterpReals4_type will be initialized and populated with mapping weights. More...
type	horizinterpreals8_type
	Nested derived type in horiz_interp_type that holds 64-bit interpolation weights, and metadata. If 64-bit grid and data arrays are provided to horiz_interp, InterphorizInterpReals8_type will be initialized and populated with mapping weights. More...
interface	stats
	Generic interfaces to compute statistics for bilinear and spherical interpolation Calls stats_r4 when the output data is in 32-bit precision. Calls stats_r8 when the output data is in 64-bit precision. More...

Functions/Subroutines
subroutine	horiz_interp_type_eq (horiz_interp_out, horiz_interp_in)
subroutine	horiz_interp_type_eq (horiz_interp_out, horiz_interp_in)
	stats_r4
	stats_r8

Variables
real(kind=r8_kind), dimension(:,:), allocatable	area_dst
	holds destination grid area.
real(kind=r4_kind), dimension(:,:), allocatable	area_dst
	holds destination grid area.
real(kind=r8_kind), dimension(:), allocatable	area_frac_dst
	holds interpolation weights for conservative interpolation, version2
real(kind=r4_kind), dimension(:), allocatable	area_frac_dst
	holds interpolation weights for conservative interpolation, version2
real(kind=r8_kind), dimension(:,:), allocatable	area_src
	holds source grid area
real(kind=r4_kind), dimension(:,:), allocatable	area_src
	holds source grid area.
integer, parameter, public	bicubic = 4
	is an internally used parameter to mark bicubic interpolation
integer, parameter, public	bilinear = 2
	is an internally used parameter to mark bilinear interpolation
real(kind=r8_kind), dimension(:,:), allocatable	faci
	holds weights for conservative interpolation version 1
real(kind=r4_kind), dimension(:,:), allocatable	faci
	holds weights for conservative interpolation version 1
real(kind=r8_kind), dimension(:,:), allocatable	facj
	holds weights for conservative interpolation version 1
real(kind=r4_kind), dimension(:,:), allocatable	facj
	holds weights for conservative interpolation version 1
logical, dimension(:,:), allocatable	found_neighbors
	is not used
type(horizinterpreals4_type)	horizinterpreals4_type
	holds more 32-bit floating point data required for interpolation.
type(horizinterpreals8_type)	horizinterpreals8_type
	holds more 64-bit floating point data required for interpolation.
logical	i_am_initialized =.false.
	is set to .true. in horiz_interp_new. Horiz_interp_base will fail if I_am_initialized = .false.
integer, dimension(:), allocatable	i_dst
	are the destination grid mapping indices in the x-direction for conservative interpolation, version 2.
integer, dimension(:,:,:), allocatable	i_lon
	contains the source grid mapping indices in x-direction for bicubic and bilinear interpolation
integer, dimension(:), allocatable	i_src
	are the source grid mapping indices in the x-direction for conservative interpolation, version 2.
integer, dimension(:,:), allocatable	ilon
	contains the source grid mapping index in x-direction for conservative interpolation, version 1
integer	interp_method
	is the interpolation method set to 1 for conservative; 2 for bilinear; 3 for spherical; 4 for bicubic.
logical	is_allocated = .false.
	is .true. if Interp is populated
logical	is_allocated = .false.
	is .true. if Interp is populated
integer, dimension(:), allocatable	j_dst
	are the destination grid mapping indices in the y-direction for conservative interpolation, version 2.
integer, dimension(:,:,:), allocatable	j_lat
	contains the source grid cell indices in y-direction for bicubic and bilinear interpolation
integer, dimension(:), allocatable	j_src
	are the source grid mapping indices in the y-direction for conservative interpolation, version 2.
integer, dimension(:,:), allocatable	jlat
	contains the source grid mapping indices in y-direction for conservative interpolation, version 1
real(kind=r8_kind), dimension(:), allocatable	lat_in
	holds the latitude coordinates on the source grid
real(kind=r4_kind), dimension(:), allocatable	lat_in
	holds the latitude coordinates on the source grid
real(kind=r8_kind), dimension(:), allocatable	lon_in
	holds the longitude coordinates on the source grid
real(kind=r4_kind), dimension(:), allocatable	lon_in
	holds the longitude coordinates on the source grid
real(kind=r8_kind), dimension(:,:), allocatable	mask_in
	masks the input grid to skip input cells when interpolation
real(kind=r4_kind), dimension(:,:), allocatable	mask_in
	masks the input grid to skip input cells when interpolation
real(kind=r8_kind)	max_src_dist
	sets to max_dist in spherical interpolation
real(kind=r4_kind)	max_src_dist
	sets to max_dist in spherical interpolation
integer	nlat_dst
	is the size of destination grid in the y direction
integer	nlat_src
	is the size of source grid in the y direction
integer	nlon_dst
	is the size of destination grid in the x direction
integer	nlon_src
	is the size of source grid in the x direction
integer, dimension(:,:), allocatable	num_found
	stores the number of neighbors found in spherical interpolation
integer	nxgrid
	is the number of exchange grid cells for conservative interpolation, version 2.
real(kind=r8_kind), dimension(:,:), allocatable	rat_x
	holds (x_dest -x_src_r)/(x_src_l -x_src_r) for bicubic interpolation
real(kind=r4_kind), dimension(:,:), allocatable	rat_x
	holds (x_dest -x_src_r)/(x_src_l -x_src_r) for bicubic interpolation
real(kind=r8_kind), dimension(:,:), allocatable	rat_y
	holds (y_src_l -y_src_r)/(y_src_l -y_src_r) for bicubic interpolation
real(kind=r4_kind), dimension(:,:), allocatable	rat_y
	holds (y_src_l -y_src_r)/(y_src_l -y_src_r) for bicubic interpolation
integer, parameter, public	spherical = 3
	is an internally used parameter to mark spherical interpolation
real(kind=r8_kind), dimension(:,:,:), allocatable	src_dist
	holds distance between destination grid and neighbor source grid in spherical interpolation.
real(kind=r4_kind), dimension(:,:,:), allocatable	src_dist
	holds distance between destination grid and neighbor source grid in spherical interpolation.
integer	version
	is set to 1 in horiz_interp_conserve_1dx1d_r4/8. Else set to 2; only used for conservative interpolation.
real(kind=r8_kind), dimension(:,:,:), allocatable	wti
	holds interpolation weights for bilinear interpolation; x-derivatives for bicubic interpolation.
real(kind=r4_kind), dimension(:,:,:), allocatable	wti
	holds interpolation weights for bilinear interpolation; x-derivatives for bicubic interpolation.
real(kind=r8_kind), dimension(:,:,:), allocatable	wtj
	holds interpolation weights for bilinear interpolation; y-derivatives for bicubic interpolation.
real(kind=r4_kind), dimension(:,:,:), allocatable	wtj
	holds interpolation weights for bilinear interpolation; y-derivatives for bicubic interpolation.

Detailed Description

Author

Zhi Liang

Defines horiz_interp_type

---

Data Type Documentation

horiz_interp_type_mod::assignment(=)

interface horiz_interp_type_mod::assignment(=)

Interface to override the "=" operator to a horiz_interp_eq that will copy the Interp instance of horiz_interp_type with Interp2 = Interp1.

Definition at line 49 of file horiz_interp_type.F90.

Private Member Functions
subroutine	horiz_interp_type_eq (horiz_interp_out, horiz_interp_in)

horiz_interp_type_mod::horiz_interp_type

type horiz_interp_type_mod::horiz_interp_type

Datatype holding interpolation weights, mapping indices, and metadata for horizontal interpolation. All real members are stored in horizInterpReals8_type if the grid and data are represented in 64-bit floating point precision or horizInterpReals4_type if the grid and data are represented in 32-bit floating point precision. Only one type, horizInterpReals4_type or horizInterpReals8_type, is allocated and used for interpolation.

Definition at line 141 of file horiz_interp_type.F90.

Collaboration diagram for horiz_interp_type:

Private Attributes
logical, dimension(:,:), allocatable	found_neighbors
	is not used
type(horizinterpreals4_type)	horizinterpreals4_type
	holds more 32-bit floating point data required for interpolation.
type(horizinterpreals8_type)	horizinterpreals8_type
	holds more 64-bit floating point data required for interpolation.
logical	i_am_initialized =.false.
	is set to .true. in horiz_interp_new. Horiz_interp_base will fail if I_am_initialized = .false.
integer, dimension(:), allocatable	i_dst
	are the destination grid mapping indices in the x-direction for conservative interpolation, version 2.
integer, dimension(:,:,:), allocatable	i_lon
	contains the source grid mapping indices in x-direction for bicubic and bilinear interpolation
integer, dimension(:), allocatable	i_src
	are the source grid mapping indices in the x-direction for conservative interpolation, version 2.
integer, dimension(:,:), allocatable	ilon
	contains the source grid mapping index in x-direction for conservative interpolation, version 1
integer	interp_method
	is the interpolation method set to 1 for conservative; 2 for bilinear; 3 for spherical; 4 for bicubic.
integer, dimension(:), allocatable	j_dst
	are the destination grid mapping indices in the y-direction for conservative interpolation, version 2.
integer, dimension(:,:,:), allocatable	j_lat
	contains the source grid cell indices in y-direction for bicubic and bilinear interpolation
integer, dimension(:), allocatable	j_src
	are the source grid mapping indices in the y-direction for conservative interpolation, version 2.
integer, dimension(:,:), allocatable	jlat
	contains the source grid mapping indices in y-direction for conservative interpolation, version 1
integer	nlat_dst
	is the size of destination grid in the y direction
integer	nlat_src
	is the size of source grid in the y direction
integer	nlon_dst
	is the size of destination grid in the x direction
integer	nlon_src
	is the size of source grid in the x direction
integer, dimension(:,:), allocatable	num_found
	stores the number of neighbors found in spherical interpolation
integer	nxgrid
	is the number of exchange grid cells for conservative interpolation, version 2.
integer	version
	is set to 1 in horiz_interp_conserve_1dx1d_r4/8. Else set to 2; only used for conservative interpolation.

horiz_interp_type_mod::horizinterpreals4_type

type horiz_interp_type_mod::horizinterpreals4_type

Nested derived type in horiz_interp_type that holds 32-bit interpolation weights, and metadata. If 32-bit grid and data arrays are provided to horiz_interp, InterphorizInterpReals4_type will be initialized and populated with mapping weights.

Definition at line 102 of file horiz_interp_type.F90.

Collaboration diagram for horizinterpreals4_type:

Private Attributes
real(kind=r4_kind), dimension(:,:), allocatable	area_dst
	holds destination grid area.
real(kind=r4_kind), dimension(:), allocatable	area_frac_dst
	holds interpolation weights for conservative interpolation, version2
real(kind=r4_kind), dimension(:,:), allocatable	area_src
	holds source grid area.
real(kind=r4_kind), dimension(:,:), allocatable	faci
	holds weights for conservative interpolation version 1
real(kind=r4_kind), dimension(:,:), allocatable	facj
	holds weights for conservative interpolation version 1
logical	is_allocated = .false.
	is .true. if Interp is populated
real(kind=r4_kind), dimension(:), allocatable	lat_in
	holds the latitude coordinates on the source grid
real(kind=r4_kind), dimension(:), allocatable	lon_in
	holds the longitude coordinates on the source grid
real(kind=r4_kind), dimension(:,:), allocatable	mask_in
	masks the input grid to skip input cells when interpolation
real(kind=r4_kind)	max_src_dist
	sets to max_dist in spherical interpolation
real(kind=r4_kind), dimension(:,:), allocatable	rat_x
	holds (x_dest -x_src_r)/(x_src_l -x_src_r) for bicubic interpolation
real(kind=r4_kind), dimension(:,:), allocatable	rat_y
	holds (y_src_l -y_src_r)/(y_src_l -y_src_r) for bicubic interpolation
real(kind=r4_kind), dimension(:,:,:), allocatable	src_dist
	holds distance between destination grid and neighbor source grid in spherical interpolation.
real(kind=r4_kind), dimension(:,:,:), allocatable	wti
	holds interpolation weights for bilinear interpolation; x-derivatives for bicubic interpolation.
real(kind=r4_kind), dimension(:,:,:), allocatable	wtj
	holds interpolation weights for bilinear interpolation; y-derivatives for bicubic interpolation.

horiz_interp_type_mod::horizinterpreals8_type

type horiz_interp_type_mod::horizinterpreals8_type

Nested derived type in horiz_interp_type that holds 64-bit interpolation weights, and metadata. If 64-bit grid and data arrays are provided to horiz_interp, InterphorizInterpReals8_type will be initialized and populated with mapping weights.

Definition at line 65 of file horiz_interp_type.F90.

Collaboration diagram for horizinterpreals8_type:

Private Attributes
real(kind=r8_kind), dimension(:,:), allocatable	area_dst
	holds destination grid area.
real(kind=r8_kind), dimension(:), allocatable	area_frac_dst
	holds interpolation weights for conservative interpolation, version2
real(kind=r8_kind), dimension(:,:), allocatable	area_src
	holds source grid area
real(kind=r8_kind), dimension(:,:), allocatable	faci
	holds weights for conservative interpolation version 1
real(kind=r8_kind), dimension(:,:), allocatable	facj
	holds weights for conservative interpolation version 1
logical	is_allocated = .false.
	is .true. if Interp is populated
real(kind=r8_kind), dimension(:), allocatable	lat_in
	holds the latitude coordinates on the source grid
real(kind=r8_kind), dimension(:), allocatable	lon_in
	holds the longitude coordinates on the source grid
real(kind=r8_kind), dimension(:,:), allocatable	mask_in
	masks the input grid to skip input cells when interpolation
real(kind=r8_kind)	max_src_dist
	sets to max_dist in spherical interpolation
real(kind=r8_kind), dimension(:,:), allocatable	rat_x
	holds (x_dest -x_src_r)/(x_src_l -x_src_r) for bicubic interpolation
real(kind=r8_kind), dimension(:,:), allocatable	rat_y
	holds (y_src_l -y_src_r)/(y_src_l -y_src_r) for bicubic interpolation
real(kind=r8_kind), dimension(:,:,:), allocatable	src_dist
	holds distance between destination grid and neighbor source grid in spherical interpolation.
real(kind=r8_kind), dimension(:,:,:), allocatable	wti
	holds interpolation weights for bilinear interpolation; x-derivatives for bicubic interpolation.
real(kind=r8_kind), dimension(:,:,:), allocatable	wtj
	holds interpolation weights for bilinear interpolation; y-derivatives for bicubic interpolation.

horiz_interp_type_mod::stats

interface horiz_interp_type_mod::stats

Generic interfaces to compute statistics for bilinear and spherical interpolation Calls stats_r4 when the output data is in 32-bit precision. Calls stats_r8 when the output data is in 64-bit precision.

Definition at line 56 of file horiz_interp_type.F90.

Private Member Functions
	stats_r4
	stats_r8

Function/Subroutine Documentation

horiz_interp_type_eq() [1/2]

subroutine horiz_interp_type_eq ( type(horiz_interp_type), intent(inout)  horiz_interp_out, type(horiz_interp_type), intent(in)  horiz_interp_in  )

private

Subroutine invoked when calling the "=" operator to copy all members of input horiz_interp_type into another instance of horiz_interp_type. Do not call subroutine directly. Instead, for copying, use the "=" operator: Interp2 = Interp1.

Parameters

[in,out]	horiz_interp_out	will contain the copied horiz_interp_type
[in]	horiz_interp_in	is the horiz_interp_type to copy

Definition at line 192 of file horiz_interp_type.F90.

horiz_interp_type_eq() [2/2]

subroutine horiz_interp_type_mod::horiz_interp_type_eq ( type(horiz_interp_type), intent(inout)  horiz_interp_out, type(horiz_interp_type), intent(in)  horiz_interp_in  )

private

Subroutine invoked when calling the "=" operator to copy all members of input horiz_interp_type into another instance of horiz_interp_type. Do not call subroutine directly. Instead, for copying, use the "=" operator: Interp2 = Interp1.

Parameters

[in,out]	horiz_interp_out	will contain the copied horiz_interp_type
[in]	horiz_interp_in	is the horiz_interp_type to copy

Definition at line 192 of file horiz_interp_type.F90.