[ncl-talk] Problem with Roms_utils.ncl

Dennis Shea shea at ucar.edu
Mon Oct 3 09:27:17 MDT 2016


Sometimes the oldest debug tools (eg: print) can be used to check an issue:

         out1   = roms_3d_interp(his1,grid,variable,rec,depth)
         printVarSummary(out1)
         printMinMax(out1,1)

         print(out1(52:65,85:105))    ; print values in region with 'blank

Variable: OUT1
Type: float
Total Size: 32292 bytes
            8073 values
Number of Dimensions: 2
Dimensions and sizes:   [69] x [117]
Coordinates:
Number Of Attributes: 1
  _FillValue :  1e+37
(0)
(0)     min=19.8764   max=26.2272


Variable: OUT1 (subsection)
Type: float
Total Size: 1176 bytes
            294 values
Number of Dimensions: 2
Dimensions and sizes:   [14] x [21]
Coordinates:
Number Of Attributes: 1
  _FillValue :  1e+37
(0,0)   25.70767
(0,1)   25.75091
(0,2)   25.79064
(0,3)   25.85184
[snip]
(2,5)   25.95704
(2,6)   25.99549
(2,7)   26.03482
(2,8)   1e+37          <=====
(2,9)   1e+37
(2,10)  1e+37
(2,11)  1e+37
(2,12)  26.1619
(2,13)  26.16949
(2,14)  26.16148
[snip]
3,4)   25.89764
(3,5)   25.95115
(3,6)   1e+37
(3,7)   1e+37
(3,8)   1e+37
(3,9)   1e+37
(3,10)  1e+37
(3,11)  1e+37
(3,12)  1e+37
(3,13)  1e+37
(3,14)  1e+37
(3,15)  1e+37
(3,16)  1e+37
(3,17)  26.16814
(3,18)  26.16628
(3,19)  26.1693
(3,20)  26.16752
(4,0)   25.71928
(4,1)   25.74613
(4,2)   25.78636
(4,3)   25.83197
(4,4)   25.89648
(4,5)   1e+37
(4,6)   1e+37
(4,7)   1e+37
(4,8)   1e+37
(4,9)   1e+37
(4,10)  1e+37
(4,11)  1e+37
(4,12)  1e+37
(4,13)  1e+37
(4,14)  1e+37
(4,15)  1e+37
(4,16)  1e+37
(4,17)  1e+37
(4,18)  1e+37
(4,19)  26.17847
(4,20)  26.19264
(5,0)   25.70962
(5,1)   25.73441
(5,2)   25.76884
(5,3)   25.8217
(5,4)   25.87821
(5,5)   1e+37
(5,6)   1e+37
(5,7)   1e+37
[SNIP]

re: "I can usually plot in other languages (matlab, for example)..."

If that region is being plotted, then Matlab is filling the region.

Good luck








On Mon, Oct 3, 2016 at 6:16 AM, <leilane.passos at inpe.br> wrote:

> Hi,
>
> I'm plotting sea surface temperature from an output of the ROMS, but a
> plot region is completely white. I checked the file and it is correct, I
> can usually plot in other languages (matlab, for example), the problem
> occurs after calling the script Roms_utils.ncl (https://www.ncl.ucar.edu/
> Applications/Scripts/ROMS_utils.ncl).
>
> I had to make 2 modifications to be able to use this script:
>
> 1 - comment some lines: 5-133 - I do not know why, but the script does
> not run if these lines are uncommented.
>
> 2 - add another parameter to call function roms_3d_interp: in my case the
> variable h (Final bathymetry at RHO-points) and the variable that will be
> interpolated (temp: time-averaged potential temperature) don't be in the
> same file. h is in my *grid file* and temp is in the *output model file*.
> So, I added a new parameter when I calling the function -  the file_grid:
>
> 137  function roms_3d_interp( file_handle, *file_grid*, var:string, \
> 138  rec:integer, z:numeric )
>
> and I changed the lines 166-176 and 208: to call the file that contains
> the variable h
>
> 166  if(typeof(file_grid).eq."file") then
> 167      ISFILE = True
> 168      nc_grid = file_grid
> 169    else if(typeof(file_grid).eq."list") then
> 170      ISFILE = False
> 171      nc_grid = file_grid[0]
> 172    else
> 173      print("roms_interp_3d: error: the first argument must be a file
> or a list of files opened with addfile or addfiles")
> 174      return
> 175    end if
> 176  end if
>
> ...
>
> 208   if(isfilevar(*nc_grid*,"h"))
>
>
>
> After this function a region with constant temperature is interpolated
> with missing values. The figures and the scripts are attached.  The files:
> nested_avg_19811216_LEILANE.nc and mesh_donor_grd_004nested_
> LEILANE.nc are in the FTP server. My NCL version is 6.3.0.
>
> I'm new in NCL and I'll appreciate any help.
>
>
> Best regards.
>
> Leilane
>
>
>
>
>
>
>
>
> _______________________________________________
> ncl-talk mailing list
> ncl-talk at ucar.edu
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20161003/077403e6/attachment.html 


More information about the ncl-talk mailing list