| 1 |
guez |
90 |
module regr_pr_int_m |
| 2 |
|
|
|
| 3 |
|
|
implicit none |
| 4 |
|
|
|
| 5 |
|
|
contains |
| 6 |
|
|
|
| 7 |
|
|
subroutine regr_pr_int(ncid, name, julien, top_value, v3) |
| 8 |
|
|
|
| 9 |
|
|
! "regr_pr_int" stands for "regrid pressure interpolation". |
| 10 |
|
|
|
| 11 |
|
|
! This procedure reads a 2D latitude-pressure field from a NetCDF |
| 12 |
|
|
! file, at a given day, regrids the field in pressure to the LMDZ |
| 13 |
|
|
! vertical grid and packs it to the LMDZ horizontal "physics" |
| 14 |
|
|
! grid. We assume that, in the input file, the field has 3 |
| 15 |
|
|
! dimensions: latitude, pressure, julian day. We assume that |
| 16 |
|
|
! latitudes are in ascending order in the input file. We assume |
| 17 |
|
|
! that the input data is already on the LMDZ "rlatu" latitude |
| 18 |
|
|
! grid. |
| 19 |
|
|
|
| 20 |
|
|
! The target vertical LMDZ grid is the grid of mid-layers. |
| 21 |
|
|
! The input data does not depend on longitude, but the pressure |
| 22 |
|
|
! at LMDZ mid-layers does. |
| 23 |
|
|
! Therefore, the values on the LMDZ grid do depend on longitude. |
| 24 |
|
|
! Regridding is by linear interpolation. |
| 25 |
|
|
|
| 26 |
|
|
use dimens_m, only: iim, jjm, llm |
| 27 |
|
|
use dimphy, only: klon |
| 28 |
|
|
use grid_change, only: dyn_phy |
| 29 |
|
|
use netcdf95, only: nf95_inq_varid, nf95_get_var |
| 30 |
|
|
use nr_util, only: assert |
| 31 |
|
|
use numer_rec_95, only: regr1_lint |
| 32 |
|
|
use press_coefoz_m, only: plev |
| 33 |
|
|
use pressure_var, only: pls |
| 34 |
|
|
|
| 35 |
|
|
integer, intent(in):: ncid ! NetCDF ID of the file |
| 36 |
|
|
character(len=*), intent(in):: name ! of the NetCDF variable |
| 37 |
|
|
integer, intent(in):: julien ! jour julien, 1 <= julien <= 360 |
| 38 |
|
|
|
| 39 |
|
|
real, intent(in):: top_value |
| 40 |
|
|
! (extra value of field at 0 pressure) |
| 41 |
|
|
|
| 42 |
|
|
real, intent(out):: v3(:, :) ! (klon, llm) |
| 43 |
|
|
! Regridded field on the partial "physics" grid. "v3(i, k)" is at |
| 44 |
|
|
! longitude "xlon(i)", latitude "xlat(i)", middle of layer "k". |
| 45 |
|
|
|
| 46 |
|
|
! Variables local to the procedure: |
| 47 |
|
|
|
| 48 |
|
|
integer varid, ncerr ! for NetCDF |
| 49 |
|
|
|
| 50 |
|
|
real v1(jjm + 1, 0:size(plev)) |
| 51 |
|
|
! Input field at day "julien". "v1(j, k >=1)" is at latitude |
| 52 |
|
|
! "rlatu(j)" and pressure "plev(k)". |
| 53 |
|
|
|
| 54 |
|
|
real v2(iim + 1, jjm + 1, llm) |
| 55 |
|
|
! Regridded field on the "dynamics" horizontal grid. "v2(i, j, k)" |
| 56 |
|
|
! is at longitude "rlonv(i)", latitude "rlatu(j)" and pressure |
| 57 |
|
|
! "pls(i, j, k)". |
| 58 |
|
|
|
| 59 |
|
|
integer i, j, k |
| 60 |
|
|
|
| 61 |
|
|
!-------------------------------------------- |
| 62 |
|
|
|
| 63 |
|
|
call assert(shape(v3) == (/klon, llm/), "regr_pr_int") |
| 64 |
|
|
|
| 65 |
|
|
call nf95_inq_varid(ncid, name, varid) |
| 66 |
|
|
|
| 67 |
|
|
! Get data at the right day from the input file: |
| 68 |
|
|
call nf95_get_var(ncid, varid, v1(:, 1:), start=(/1, 1, julien/)) |
| 69 |
|
|
! Latitudes are in ascending order in the input file while |
| 70 |
|
|
! "rlatu" is in descending order so we need to invert order: |
| 71 |
|
|
v1(:, 1:) = v1(jjm+1:1:-1, 1:) |
| 72 |
|
|
|
| 73 |
|
|
! Complete "v1" with the value at 0 pressure: |
| 74 |
|
|
v1(:, 0) = top_value |
| 75 |
|
|
|
| 76 |
|
|
! Regrid in pressure at each horizontal position: |
| 77 |
|
|
do j = 1, jjm + 1 |
| 78 |
|
|
do i = 1, iim |
| 79 |
|
|
if (dyn_phy(i, j)) then |
| 80 |
|
|
v2(i, j, llm:1:-1) & |
| 81 |
|
|
= regr1_lint(v1(j, :), (/0., plev/), pls(i, j, llm:1:-1)) |
| 82 |
|
|
! (invert order of indices because "pls" is in descending order) |
| 83 |
|
|
end if |
| 84 |
|
|
end do |
| 85 |
|
|
end do |
| 86 |
|
|
|
| 87 |
|
|
forall (k = 1:llm) v3(:, k) = pack(v2(:, :, k), dyn_phy) |
| 88 |
|
|
|
| 89 |
|
|
end subroutine regr_pr_int |
| 90 |
|
|
|
| 91 |
|
|
end module regr_pr_int_m |
Parent Directory
| 
Revision Log