WAVEWATCH III ERROR IN W3BOUNC

[Cesar-IIUNAM]

Hi,
I have generated some spectral points from a "parent" model, and I am trying to impose them as OBCs for the northern boundary of a subdomain (I previously ran this "child" model forced only with winds without problems). However, I am getting the following error:

[ctorresm@master gdm_nest_L1]$ ./ww3_bounc

              * WAVEWATCH III Bounday input prep. *  
           ===============================================

Grid name : TEST GRID (GULF OF MEXICO)

Comment character is '$'

Opening file:ww3.850W230N_201701_spec.nc
Opening file:ww3.855W230N_201701_spec.nc
Opening file:ww3.865W230N_201701_spec.nc
Opening file:ww3.875W230N_201701_spec.nc
Opening file:ww3.885W230N_201701_spec.nc
Opening file:ww3.895W230N_201701_spec.nc
Opening file:ww3.905W230N_201701_spec.nc
Opening file:ww3.915W230N_201701_spec.nc
Opening file:ww3.920W230N_201701_spec.nc

• WAVEWATCH III ERROR IN W3BOUNC:
  ILLEGAL NK, NK = 19 DIFFERS FROM NKI = 34
  IT WILL BE MANAGED SOON BY SPCONV

I converted each spectral point into individual ".nc" files, using the program "ww3_ounp."
Inside my child model, I have "ww3_bounc.inp" configured as follows:

$ WAVEWATCH III NetCDF boundary input processing $
$--------------------------------------------------------------------- $
$
$ Boundary option: READ or WRITE
$
WRITE
$
$ Interpolation method: 1: nearest
$ 2: linear interpolation
2
$ Verbose (0, 1, 2)
1
$
$ List of spectra files. These NetCDF files use the WAVEWATCH III
$ format as described in the ww3_ounp.inp file. The files are
$ defined relative to the directory in which the program is run.
$
ww3.850W230N_201701_spec.nc
ww3.855W230N_201701_spec.nc
ww3.865W230N_201701_spec.nc
ww3.875W230N_201701_spec.nc
ww3.885W230N_201701_spec.nc
ww3.895W230N_201701_spec.nc
ww3.905W230N_201701_spec.nc
ww3.915W230N_201701_spec.nc
ww3.920W230N_201701_spec.nc
'STOPSTRING'
$
$ -------------------------------------------------------------------- $
$ End of input file $
$ -------------------------------------------------------------------- $

In "ww3_grid.inp" of the child model I defined the north border as open, for that, in ww3_grid.inp I typed:

$ Input boundary points from segment data ( FROM = PART ) ------------ $
$

1 81 F
141 81 T (esto lo defino según las dimensiones de mi malla)

I would be very grateful for your help to understand the origin of the error and to be able to give a solution.

Best regards,
César.

[aliabdolali]

are spectral resolutions of parent and child setup the same???

[Cesar-IIUNAM]

Hello, @aliabdolali ..
thank you for your quick response.

I don't know if I understand your question. The domains have different spatial resolutions, but I think that's not what you mean.
Although I did the following: the parent model has a spatial resolution of 20 km and in "ww3_shel.inp" I defined spectral points every one degree, spatially distributed over what would be the northern border of my subdomain; which has a resolution of 5 km.

Please let me know if you need more information.

César.

[aliabdolali]

I meant the number of frequencies and directions, frequency increment and first frequency for both setup. They are defined in your ww3_grid.nml(inp)

[Cesar-IIUNAM]

Okay, thank you very much.
They are not the same.
For the father I have defined:
1.1 0.041 34 72 0.5
For the son:
1.1 0.041 19 24 0.5

Ok, it seems that the problem is in the frequency number between the two models? 19 and 34 and that's why the message "ILLEGAL NK, NK = 19 DIFFERS FROM NKI = 34" ?

[aliabdolali]

exactly :) this is the problem.

[aliabdolali]

ok, is this difference intentional? I mean in the parern domain, you are considering frequency range of 0.041-0.95 Hz with directional resolution of 5 degrees.
in the child domain, you just consider f=0.041-0.22 Hz with directional resolution of 15 degrees

I can give you pre-processing scripts which does the conversion, if you really want to interpolate.