[ncl-install] plot

fereshteh komijani fereshtehkomijani at gmail.com
Sat Oct 19 04:38:05 MDT 2013


Dear all

Briefly I use ROMS model. In this model there are masking process to
separate land and sea. One of its outputs is ocean_history.nc file. Its
header contain

*Code:*

float u(ocean_time, s_rho, eta_u, xi_u) ;
      u:long_name = "time-averaged u-momentum component" ;
      u:units = "meter second-1" ;
      u:time = "ocean_time" ;
      u:coordinates = "lon_u lat_u s_rho ocean_time" ;
      u:field = "u-velocity, scalar, series" ;
      u:_FillValue = 1.e+37f ;

for all variables. As you can see there is _FillValue option with  1.e+37fvalue.

For plotting, there are csec.in and ccent.in in ROMS plotting package for
cross-section and countor plotting, respectively by

 $ ccntG<ccnt_test.in

$ csecG<csec_test.in

That base on ROMS manual, they plotted with ncarg.

Recently I face with one problem that makes me confused. My problem is that
when I plot some variable cross section, I see that its color bar and also
max and min values which written in plot's corner are not correct. For
example for u component max and min values are 1.83 and -2.47,
respectively. But in plot it shows values in order e+38 and whole of plot
covered by same color (attached plot). It seems that values multiplied by
one factor. I send this error to ROMS forum several times. They reply:


1-    1- Very simple. You need to pay attention to the statistics that are
dumped to standard output when plotting. Also, see the minimum and maximum
values in the bottom left corner of the plot. Notice also at the bottom of
the colorbar there is a scale x10^38. The maximum value in the plot is
around 1E+38 [image: :!:] The special value in ROMS for masking is 1.0E+37.
The plotting package is able to recognize such masked values. The *
_FillValue* is added in masked regions and it is easy recognizable by any
third party plotting software. You can try plotting your fields with any
NetCDF file tools out there



2-     2-The masked values are not NaN, but rather some strange 1.e38 value
picked to never be a valid value. You need to tell your plotting software
to use a mask for anything equal to the _FillValue. ncdump and ncview know
to do that.


3-Your choices are to look at the source code and fix it yourself or to
submit a bug report to the maker of the software.


Bas on this replies would you please help and guide e how I can do that
till be able to plot correctly.

Thanks in advance for your attention

Cheers

fereshte





-- 
***Angel***

**
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/ncl-install/attachments/20131019/9210c115/attachment-0001.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: total u velocity.png
Type: image/png
Size: 26200 bytes
Desc: not available
Url : http://mailman.ucar.edu/pipermail/ncl-install/attachments/20131019/9210c115/attachment-0001.png 


More information about the ncl-install mailing list