Just to follow on here... Fourier analysis always assumes cyclic boundary
conditions. We get around that for frequency by tapering the beginning and
end to zero (thus making it cyclic).
As Dennis suggests, one option is to pad the data to make it global.
Another would be to taper the edges and do the spectrum manually with the
function fft2df. Either route will take quite a bit of work.
Carl
On Wed, Dec 21, 2011 at 10:17 AM, Dennis Shea <shea@ucar.edu> wrote:
> The original W-K fortran code is hard-wired for longitudinally global
> data. To use the current code, you would have to pad out your data
> in some intelligent way so they are global.
>
> The wave numbers are based on this assumption.
>
> On 12/21/11 5:37 AM, Abhisek Chatterjee wrote:
> > Dear NCL users,
> >
> > I am a new user of NCL. I am trying to do a frequency_wavenumber spectra
> > as Wheeler and Kiladis (1999) for meridional velocity component of ocean
> > in Equatorial Indian Ocean.
> > But as documented in NCL user guide the input file should be cyclic in
> > longitude for this application , unfortunately it is not true in my
> > case. Please suggest me, how to do this for a particular region
> > (instead of whole globe).
> >
> > Abhisek Chatterjee
> > SRF
> > National Institute of Oceanography
> > Goa, India
> >
> >
> >
> > _______________________________________________
> > 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
>
-- Carl J. Schreck III, PhD Postdoctoral Research Associate Cooperative Institute for Climate and Satellites (CICS-NC) North Carolina State University and NOAA's National Climatic Data Center 151 Patton Avenue Asheville, NC 28801 Tel: 828-257-3140 carl@cicsnc.org http://monitor.cicsnc.org/mjo/ http://www.atmos.albany.edu/student/carl/
_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Wed Dec 21 08:53:32 2011
This archive was generated by hypermail 2.1.8 : Wed Dec 21 2011 - 10:44:06 MST