Re: NCL and OPeNDAP

Hi Brandon,

What I mean by "invalid" is that your u and v are on different lat/lon grids, but the way I cheated was by using u's lat/lon coordinate arrays for v. This means that v will not be positioned quite right.
I did this only so I could see if u and v contained valid values. How did you get u and v to be on the same lat/lon grid?

If you could provide me with your script, then I can try to see why the OPeNDAP version is not working.

--Mary

On Jul 5, 2012, at 7:10 AM, Brandon Capasso wrote:

> Mary,
>
> Thank you for your prompt reply. I have already accounted for the staggered grid in my script, I was getting blank u,v plots from the remote data due to an error with res@vfXArray/res@vfYArray. I think I was also referring to resV@sfXArray and resV@sfYArray with the same variable definitions.
>
> BUT, I'm still not entirely sure why my script was working perfectly on the netcdf file on my local machine (it was plotting the same thing as when operating on the data remotely with opendap when I removed that second definition, only defining vfXArray/vfYArray).
>
> Also, I'm not sure what you mean by saying that it's not a valid plot. I understand that the grid is curvilinear and very non-uniform, but my data plot matches the grid definition at http://tidesandcurrents.noaa.gov/ofs/cbofs/cbofs_info.html and the NOAA forecast current images on the home page.
>
> How would change your script to create a valid plot from the data? It seems to me that it's doing what it's supposed to be doing.
>
> Thanks again,
>
> Brandon
>
> On Tue, Jul 3, 2012 at 11:32 AM, Mary Haley <haley@ucar.edu> wrote:
> Brandon,
>
> Which variables are you trying to plot a vector plot of?
>
> I was able to plot "u" and "v", but I had to cheat because they weren't on the same grid.
>
> You will need to do this before you can plot u and v, since they are on staggered grids.
>
> The way I cheated was to just select a subset of u and v, such that they are the same size.
> I also had to subscript lat_u and lon_u the same way. Note that this won't create a correct
> plot, but I just wanted to verify that there were some values.
>
> See the attached script.
>
>
> --Mary
>
> On Jul 3, 2012, at 12:53 AM, Brandon Capasso wrote:
>
> > Greetings,
> >
> > I am trying to read and plot fields from a remote NetCDF file, but I am coming up with "blank" fields.
> >
> > When I use:
> >
> > cdf_file = addfile(path,"r")
> >
> > where path points to http://opendap.co-ops.nos.noaa.gov/thredds/dodsC/NOAA/CBOFS/MODELS/201207/nos.cbofs.fields.nowcast.20120703.t00z.nc, my script runs without error and apparently reads the file, but plots blank vector fields (as if there is no data). Furthermore, printVarSummary on the remote file returns the correct dimension/attribute info for the variables that I'm interested in (so I know ncl is reading the remote file).
> >
> > When operating on the same file downloaded/saved on my local machine, my script works perfectly and plots the vector fields without issue, so why am I getting blank fields when operating on the remote file?
> >
> > Regards,
> >
> > Brandon Capasso
> >
> >
> > _______________________________________________
> > ncl-talk mailing list
> > List instructions, subscriber options, unsubscribe:
> > http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
>
>
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Fri Jul 6 08:16:07 2012