Skip to content

Commit

Permalink
Fix indexing bugs in get_field_nc (#334)
Browse files Browse the repository at this point in the history
Fixed two horizontal indexing bugs in `get_field_nc`, where the difference
between the array starting index (always 1 in this subroutine) and the values
in the handle argument were not being taken into account when the array was
being passed in with only its computational domain.

Also initialized the internal `unlim_index` array in `get_netcdf_fields` to fix
a problem with using an uninitialized array that was being flagged when run in
debug mode.

With this commit, the model is once again reproducing the expected answers when
rescaling is applied for vertical distances or time.

Co-authored-by: Marshall Ward <marshall.ward@gmail.com>
  • Loading branch information
Hallberg-NOAA and marshallward authored Mar 26, 2023
1 parent ceb4c92 commit 37389b5
Show file tree
Hide file tree
Showing 3 changed files with 39 additions and 21 deletions.
9 changes: 5 additions & 4 deletions src/framework/MOM_io.F90
Original file line number Diff line number Diff line change
Expand Up @@ -2057,23 +2057,24 @@ subroutine MOM_read_data_2d(filename, fieldname, data, MOM_Domain, &
end subroutine MOM_read_data_2d


!> Read a 2d array from file using native netCDF I/O.
!> Read a 2d array (which might have halos) from a file using native netCDF I/O.
subroutine read_netCDF_data_2d(filename, fieldname, values, MOM_Domain, &
timelevel, position, rescale)
character(len=*), intent(in) :: filename
!< Input filename
character(len=*), intent(in) :: fieldname
!< Field variable name
real, intent(out) :: values(:,:)
!< Field value
real, intent(inout) :: values(:,:)
!< Field values read from the file. It would be intent(out) but for the
!! need to preserve any initialized values in the halo regions.
type(MOM_domain_type), intent(in) :: MOM_Domain
!< Model domain decomposition
integer, optional, intent(in) :: timelevel
!< Time level to read in file
integer, optional, intent(in) :: position
!< Grid positioning flag
real, optional, intent(in) :: rescale
!< Rescale factor
!< Rescale factor, omitting this is the same as setting it to 1.

integer :: turns
! Number of quarter-turns from input to model grid
Expand Down
48 changes: 31 additions & 17 deletions src/framework/MOM_io_file.F90
Original file line number Diff line number Diff line change
Expand Up @@ -1681,15 +1681,18 @@ subroutine read_field_chksum_nc(handle, field, chksum, valid_chksum)
end subroutine read_field_chksum_nc


!> Read the values of a netCDF field
!> Read the values of a netCDF field into an array that might have halos
subroutine get_field_nc(handle, label, values, rescale)
class(MOM_netcdf_file), intent(in) :: handle
! Handle of netCDF file to be read
!< Handle of netCDF file to be read
character(len=*), intent(in) :: label
! Field variable name
real, intent(out) :: values(:,:)
! Field values read from file
!< Field variable name
real, intent(inout) :: values(:,:)
!< Field values read from the file. It would be intent(out) but for the
!! need to preserve any initialized values in the halo regions.
real, optional, intent(in) :: rescale
!< A multiplicative rescaling factor for the values that are read.
!! Omitting this is the same as setting it to 1.

logical :: data_domain
! True if values matches the data domain size
Expand All @@ -1701,10 +1704,12 @@ subroutine get_field_nc(handle, label, values, rescale)
! Index bounds of compute domain
integer :: isd, ied, jsd, jed
! Index bounds of data domain
integer :: iscl, iecl, jscl, jecl
! Local 1-based index bounds of compute domain
integer :: bounds(2,2)
! Index bounds of domain
real, allocatable :: values_nc(:,:)
! Local copy of field, used for data-decomposed I/O
real, allocatable :: values_c(:,:)
! Field values on the compute domain, used for copying to a data domain

isc = handle%HI%isc
iec = handle%HI%iec
Expand All @@ -1727,36 +1732,45 @@ subroutine get_field_nc(handle, label, values, rescale)

field_nc = handle%fields%get(label)

if (data_domain) then
allocate(values_nc(isc:iec,jsc:jec))
values(:,:) = 0.
endif
if (data_domain) &
allocate(values_c(1:iec-isc+1,1:jec-jsc+1))

if (handle%domain_decomposed) then
bounds(1,:) = [isc, jsc] + [handle%HI%idg_offset, handle%HI%jdg_offset]
bounds(2,:) = [iec, jec] + [handle%HI%idg_offset, handle%HI%jdg_offset]
if (data_domain) then
call read_netcdf_field(handle%handle_nc, field_nc, values_nc, bounds=bounds)
call read_netcdf_field(handle%handle_nc, field_nc, values_c, bounds=bounds)
else
call read_netcdf_field(handle%handle_nc, field_nc, values, bounds=bounds)
endif
else
if (data_domain) then
call read_netcdf_field(handle%handle_nc, field_nc, values_nc)
call read_netcdf_field(handle%handle_nc, field_nc, values_c)
else
call read_netcdf_field(handle%handle_nc, field_nc, values)
endif
endif

if (data_domain) &
values(isc:iec,jsc:jec) = values_nc(:,:)
if (data_domain) then
iscl = isc - isd + 1
iecl = iec - isd + 1
jscl = jsc - jsd + 1
jecl = jec - jsd + 1

values(iscl:iecl,jscl:jecl) = values_c(:,:)
else
iscl = 1
iecl = iec - isc + 1
jscl = 1
jecl = jec - jsc + 1
endif

! NOTE: It is more efficient to do the rescale in-place while copying
! values_nc(:,:) to values(:,:). But since rescale is only present for
! values_c(:,:) to values(:,:). But since rescale is only present for
! debugging, we can probably disregard this impact on performance.
if (present(rescale)) then
if (rescale /= 1.0) then
values(isc:iec,jsc:jec) = rescale * values(isc:iec,jsc:jec)
values(iscl:iecl,jscl:jecl) = rescale * values(iscl:iecl,jscl:jecl)
endif
endif
end subroutine get_field_nc
Expand Down
3 changes: 3 additions & 0 deletions src/framework/MOM_netcdf.F90
Original file line number Diff line number Diff line change
Expand Up @@ -665,6 +665,9 @@ subroutine get_netcdf_fields(handle, axes, fields)
call check_netcdf_call(rc, 'get_netcdf_fields', &
'File "' // trim(handle%filename) // '"')

! Initialize unlim_index with an unreachable value (outside [1,ndims])
unlim_index = -1

allocate(axes(ndims))
do i = 1, ndims
rc = nf90_inquire_dimension(handle%ncid, dimids(i), name=label, len=len)
Expand Down

0 comments on commit 37389b5

Please sign in to comment.