Re: Regridding Daymet Joined Tiles with ESMF_Regridding on NCL 6.1.0

From: Mary Haley <haley_at_nyahnyahspammersnyahnyah>
Date: Mon Nov 05 2012 - 08:46:31 MST

Ping,

In order to help you, we have to see your full script and all your data files. Without knowing what your variables look like, we can't tell you how to correctly do the regridding.

You can put your data and script(s) on our anonymous ftp. Be sure to include *all* data files and scripts needed to run your main script.

Here are the anonymous ftp instructions:

http://www.ncl.ucar.edu/report_bug.shtml#HowToFTP

Thanks,

--Mary

On Nov 3, 2012, at 9:37 PM, Ping Yang wrote:

> Dear NCL,
>
> I am still stuck at here for the joined Daymet grid.
>
> On Tue, Oct 30, 2012 at 6:40 PM, Dave Allured <dave.allured@noaa.gov> wrote:
> Ping,
>
> Your script is using sfile->lat and sfile->lon. If this is the same
> merged file that you showed yesterday, then sfile->lat and sfile->lon
> will contain gaps filled with zeros rather than valid lats and lons.
> This may or may not cause the regrid function to fail. Perhaps
> someone familiar with that function could say whether this would work.
> the zeros value made the ESMF regriding fail.
>
> The simple remedy would be to operate on a rectangular subset of the
> merged file that does not have any coordinate gaps.
> What is a rectangular subset? I tried to use cdo to sellonlatbox to subset but it told me the grid is a generic grid which can not be subseted.
> The better remedy
> would be to fill in all of the missing lats and lons, which is beyond
> the scope of the current version of my join script.
> I am thinking is that possible to interpolate those missing lats and lons instead of put zero value to them. Do you have some idea?
>
> Another solution would be to have the regrid function compute source
> lats and lons from native grid projection parameters, rather than the
> source 2-D coordinates. I do not know whether *any* NCL regrid
> functions have this kind of support.
> I hope if there is an ESMF specialist can give some advice on this.
>
> Thank you so much.
>
> Regards,
>
> Ping
> _______________________________________________
> 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 Mon Nov 5 08:46:41 2012

This archive was generated by hypermail 2.1.8 : Tue Nov 06 2012 - 15:05:49 MST