[2] There is no lowess filter built into in NCL.
Most commonly a weighted running average is used.
[1] For Taylor diagrams
( http://www.ncl.ucar.edu/Applications/taylor.shtml )
you need:
(a) (centered) pattern correlations (also called cross correlations)
http://www.ncl.ucar.edu/Document/Functions/Contributed/pattern_cor.shtml
(b) ratios of areally weighted variances
(c) % bias. This is an extion to the classic Taylor diagram.
See Example 8.
A sample processing script is Example 7. Note: this was written before
the 'pattern_cor. The section on computing areally weight variances
wgt = conform(vcntl, wt , rank-2)
;;wgt = mask(wgt, lsflag.eq.0, False) ; if desired
;;wgt = mask(wgt, lsflag.eq.1, False)
; temporary variables
sumw = sum(wgt)
sumwc = sum(wgt*vcntl)
sumwt = sum(wgt*vtest)
; wgted areal mean
wmean_cntl = sumwc/sumw
wmean_test = sumwt/sumw
; wgted areal variance
wvar_cntl = sum(wgt*(vcntl-wmean_cntl)^2)/sumw
wvar_test = sum(wgt*(vtest-wmean_test)^2)/sumw
ratio(nc,nv) = (wvar_test/wvar_cntl)^0.5
Good luck
On 6/29/12 1:57 PM, Noel Aloysius wrote:
> Hi NCL,
>
> I would like to hear your suggestions/directions on the following,
>
> 1) I want to create a Taylor diagram comparing GCM performance (10 GCMs)
> of three variables at three time intervals - early, mid and late 20th
> century. Are there NCL examples that guide one through the input
> preparation for Taylor diagram?
>
> 2) Is there an equivalent of lowess filter type function in NCL for
> monthly time series smoothing.
>
> Thanks,
> Noel
>
>
>
>
> _______________________________________________
> 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 Jun 29 14:17:08 2012
This archive was generated by hypermail 2.1.8 : Fri Jun 29 2012 - 15:46:13 MDT