| Re: wavenum, frequency FFT plot [message #66172] |
Thu, 23 April 2009 15:03  |
R.G. Stockwell
Messages: 363
Registered: July 1999
|
Senior Member |
|
|
<kishore1818@gmail.com> wrote in message
news:c004f032-02b9-4990-a268-73966d3c6383@a5g2000pre.googleg roups.com...
> Hi,
> I have time vs latitudinal values and I would like to estimate the
> power spectral density 2-D do the 2-D FFT to get the frequency VS wave
> number .
> Example : data sets are (30,8); 30=days values and 8 latitudinal avg.
> data
> I am planning to wave number, frequency and power contour plot.
> x-axis zonal wave number range is -7 to 7
> Y-axis frequency range is -1. To 1.
> Does anybody have a good example program on 2D FFT in IDL program.
>
> Thanking you,
>
> Kishore
You mean longitudinal values (for zonal wavenumber)?
And is it global (45 deg sampling)?
just call FFT on the data, keep in mind that it is symmetrical
across the origin, and that you're index numbers are the wavenumbers
in the k_lon direction.
The sampling in time is i/(NT), where T is the sampling interval in time.
cheers,
bob
|
|
|
|
| Re: wavenum, frequency FFT plot [message #66239 is a reply to message #66172] |
Mon, 27 April 2009 14:28  |
kishore1818
Messages: 24
Registered: June 2007
|
Junior Member |
|
|
On Apr 23, 3:03 pm, "R.G. Stockwell" <noemai...@please.com> wrote:
> <kishore1...@gmail.com> wrote in message
>
> news:c004f032-02b9-4990-a268-73966d3c6383@a5g2000pre.googleg roups.com...
>
>> Hi,
>> I have time vs latitudinal values and I would like to estimate the
>> power spectral density 2-D do the 2-D FFT to get the frequency VS wave
>> number .
>> Example : data sets are (30,8); 30=days values and 8 latitudinal avg.
>> data
>> I am planning to wave number, frequency and power contour plot.
>> x-axis zonal wave number range is -7 to 7
>> Y-axis frequency range is -1. To 1.
>> Does anybody have a good example program on 2D FFT in IDL program.
>
>> Thanking you,
>
>> Kishore
>
> You mean longitudinal values (for zonal wavenumber)?
> And is it global (45 deg sampling)?
>
> just call FFT on the data, keep in mind that it is symmetrical
> across the origin, and that you're index numbers are the wavenumbers
> in the k_lon direction.
>
> The sampling in time is i/(NT), where T is the sampling interval in time.
>
> cheers,
> bob
Hi Bob,
Thanks for your e-mail message.
Actually I have 365 days daily avg values, every 10 deg latitudinal
averages.
ie., (365,18). I need to calculate the psd for 30days data sets for
all latitudinal values.
But I am planning to plot contour: wvae no. (x axis), frequency (y-
axis) and z is power.
could you give a example program then it is very easy to understand to
me and also I am beginner in IDL.
Thanking you,
Kishore
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
