Re: IPO index calculation

Hi,
Thanks for your response. I tried to code this up (please see below). I
would like to do a Chebyshev filter with a half-power period of 11 years
(44 seasons). So, I used nwt= 45 and fca =1/44. I still can't reproduce the
values that I needed. I am not sure if I am doing anything wrong in using
the filter.
Thanks,
Mahesh

;===========================================

; Low-pass filtering

;-----------------------------

;weights for low pass filter (11 years= 44 seasons)

 nwt = 45 ; 22 seasons each end.

 fca = 1./44. ; 11-year i.e. 44 seasons

 ihp = 0 ;low_pass

 nsigma = 1. ;Lanczos sigma

 opt=0

 wgtlp = filwgts_lanczos (nwt, ihp, fca, -999., nsigma)

 SSTt_anom_lp = wgt_runave(data3d(lat|:,lon|:,time|:), wgtlp, opt)

 ;return

;----------------------------

On Thu, Apr 3, 2014 at 8:43 AM, Dennis Shea <shea@ucar.edu> wrote:

>
> It is the response function for the filter being used
> that is important. eg: See attached
>
> The current version of NCL has Lanczos filtering.
>
>
> http://www.ncl.ucar.edu/Document/Functions/Built-in/filwgts_lanczos.shtml
>
>
> See the Filter page
>
> http://www.ncl.ucar.edu/Applications/filter.shtml
>
> See Example 8. You can download and run the script to determine
> the number of weights you need for your application. You will
> lose nwgt/2 points at each end of the series.
>
> ---
>
> The 6.3.0 release will have a Butterworth filter optimized for
> a narrow passband width. Note: The Butterworth and Chebyshev filters have
> 'end effects' also.
>
>
>
>
> On 4/2/14, 2:56 PM, Mahesh Varma wrote:
>
>> Hi,
>> I am trying to calculate Interdecadal Pacific Oscillation Index using
>> HadSST2 data. I am following Parker et al., 2007 method to calculate the
>> IPO. Parker et al. use a filter called Chebyshev filter. I am not sure if
>> NCL has any built in function that does Chebyshev filtering. I could
>> calculate IPO index without the filter and but am particularly interested
>> in reproducing the monthly principal component values as in Parker et al.
>> 2007. Does any one have a script already written to calculate IPO? Any
>> help
>> will be appreciated.
>> Thanks,
>> Mahesh
>>
>>
>>
>> _______________________________________________
>> 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 Thu Apr 03 18:20:46 2014