Re: I have a question about csa2ld function.

Hi Miae,

Not all functions automatically deal with missing data, and csa2ld is one of them. One thing you could try is removing all the missing values before you interpolate.

ii = ind(.not(ismissing(xi)).and..not.ismissing(yi)))
xi_nomsg = xi(ii)
yi_nomsg = yi(ii)

Have you considered trying the ESMF regridding routines? This gives you a number of interpolation options, and does handle missing values.

You can see some examples at:

http://www.ncl.ucar.edu/Applications/ESMF.shtml

It looks like you have an unstructured grid, so search for the word “unstructured� in this page for some examples.

If you have any further questions, please post them back to ncl-talk.

Thanks,

—Mary

On Apr 11, 2014, at 10:41 AM, 김미애 <toa0710@naver.com> wrote:

> Dear NCL team,
>
> I have a question about csa2ld function, which makes interpolation with cubic spline algorithm.
> I attached two figures. First one is Arctic elevation scattered data before interpoation. Empty region (white) above N65 in first figure that has fill value 1.844674407370955e+19. Second one is a result that I did csa2ld. This looks very weird.
> I think 1.844674407370955e+19 fill values used in running csa2ld, so values are so high and looks circular shape. I thought fill value automatically is excluded with respect to function of csa2ld like a function of dim_avg_n_Wrap.
> I want to use this function without fill value.
> Are there another options or other functions to running cubic spline interpolation algorithm without fill value?
>
> <first.jpg>
> <First figure>
>
> <second.jpg>
> <Second figure>
>
>
> knots = (/4,4/)
> h1 = csa2ld(coord(:,0),coord(:,1),k1,knots,coord(:,0),coord(:,1))
>
> Sincerely yours
>
> Thank you.
>
> Miae
>
> -------------------------------------------------
> Miae Kim
> School of Urban and Environment Engineering
> Ulsan National Institute of Science and Technology (UNIST)
> UNIST-gil 50, Ulsan 689-798, South Korea
> Mobile: +82-10-9421-8287
> E-mail: toa0710@naver.com
> _______________________________________________
> 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 Tue Apr 15 20:00:57 2014