topography following coordinate

Added by Johannes Röhrs about 13 years ago

Does anyone know if it's possible to use CDO for a vertical interpolation of 3d data to stretched topography following coordinates?

I found an example to describe z levels:
https://code.zmaw.de/embedded/cdo/1.4.7/cdo.html#x1-210001.4
However, the topography following coordinate transformation depends on x and y, and I didn't see how I could specify this to the cdo's. I could either supply the transfrom equation and a topography file or alternatively a 3d mapping array.

Did anyone before me use CDO to work with ROMS data or any other terrain-following model?

Replies (3)

RE: topography following coordinate - Added by Ralf Mueller about 13 years ago

Hi Johannes!
At first, I have to get clear, what kind of data you're handling with: You have variable (let's call it) "height" witch depends on x,y and z. And the values should describe the z-coordinate values at location x,y at level z. Did I get it right?

RE: topography following coordinate - Added by Johannes Röhrs about 12 years ago

Hi again!

More than one year after I opened this thread, I noticed that since last summer, cdo (version 1.5.1) has a function available for arbitrary vertical interpolations:

cdo intlevel3d,icoordinate ifile1 ifile2 ofile

Now I tried to use intlevel3d to vertically interpolate a 3d field with vertical coordinates (z) in 'depth below sea level' to a topography following vertical coordinate (s) that depends on x and y: s=s(z,y,x). Unfortunately, cdo crashed with a 'segmentation fault'.

I tried:

> cdo intlevel3d,"input_z_grid.nc" my_variable.nc target_s_grid.nc output.nc
Segmentation fault

my input file looks like the following. my_variable.nc contains the field that is supposed to be interpolated from z,y,x to s,y,x:

cdo sinfo my_variable.nc 
   File format: netCDF
    -1 : Institut Source   Param       Time Typ  Grid Size Num  Levels Num
     1 : unknown  unknown  -1          con  F32    554106   1      35   1
   Horizontal grids :
     1 : curvilinear  > size      : dim = 554106  nx = 1106  ny = 501
                        lon_rho   : min = 2.28063246  max = 20.5218613  degree_east
                        lat_rho   : min = 61.8464315  max = 70.7475513  degree_north
                        available : xvals yvals
   Vertical grids :
     1 : generic            level : 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 
                                    19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 
                                    34 35 
cdo sinfo: Processed 1 variable. ( 0.05s )

the input grid:

cdo sinfo input_z_grid.nc 
   File format: netCDF
    -1 : Institut Source   Param       Time Typ  Grid Size Num  Levels Num
     1 : unknown  unknown  -1          con  F32    554106   1       4   1
   Horizontal grids :
     1 : curvilinear  > size      : dim = 554106  nx = 1106  ny = 501
                        lon_rho   : min = 2.28063246  max = 20.5218613  degree_east
                        lat_rho   : min = 61.8464315  max = 70.7475513  degree_north
                        available : xvals yvals
   Vertical grids :
     1 : depth_below_sea        m : 0 10 50 5000

the target grid:

cdo sinfo target_s_grid.nc 
   File format: netCDF
    -1 : Institut Source   Param       Time Typ  Grid Size Num  Levels Num
     1 : unknown  unknown  -1          con  F32    554106   1      35   1
   Horizontal grids :
     1 : curvilinear  > size      : dim = 554106  nx = 1106  ny = 501
                        lon_rho   : min = 2.28063246  max = 20.5218613  degree_east
                        lat_rho   : min = 61.8464315  max = 70.7475513  degree_north
                        available : xvals yvals
   Vertical grids :
     1 : generic            level : 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 
                                    19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 
                                    34 35 
cdo sinfo: Processed 1 variable. ( 0.04s )

This thread is related to Scott's, who is working on the same problem (to vertically interpolate data for the topography following ocean model ROMS):
https://code.zmaw.de/boards/1/topics/913#message-972

Can anyone help out here? I am not sure if I am using the interpolation function the right way, and it's difficult to find out since I don't get any output. I tried to run the example on a different computer, as well as with a different cdo version, but it does not seem to be related to my installation of cdo.

RE: topography following coordinate - Added by Scott Rowe about 12 years ago

Hi Johannes,

After running into the same problem and inspecting the code I came to the conclusion that the intlevel3d function does not handle curvilinear grids. You must use a lonlat horizontal grid type. To do this, I had to modify my lon_rho and lat_rho variables so that they were one dimensional only. CDO then interprets the grid as being lonlat and the segmentation fault is avoided.

(1-3/3)

Project

General

Profile

CDO