[NARCCAP-discuss] Common Grid for NARCCAP?

Seth McGinnis mcginnis at ucar.edu
Fri Jul 19 11:55:12 MDT 2013


Hi Ethan,

Yup, that's basically it. Whatever lat/lon values you have in the grid file,
it'll interpolate the data to those locations.

(Technically it's not working on just the nearest neighbors -- it's actually
doing fancy statistics on the entire field to take into account the spatial
correlation length and so on.  This is what makes it slower than something like
griddata() in IDL or rmc2rgrid() in NCL.)

I can't remember what all assumptions I made about the various input files, so
if it seems like the code is being finicky about your gridfile, email me
directly and I'll help you get it sorted out.

Cheers,

--Seth


On Thu, 18 Jul 2013 11:40:51 -0400
 Ethan Bodnaruk <ebodnaruk at gmail.com> wrote:
>Hi Seth,
>
>Thanks so much for your help.  I'm just starting to take a look at your
>code.  Originally I imagined a situation in which I have a location of
>interest surrounded by 4 actual domain coordinates or grids.  A program
>could take those 4 locations and interpolate to create data corresponding
>to the desired location.
>
>It looks like your program requires one input file plus the gridfile,
>though.  So if I only want to regrid for one particular location (i.e. a
>certain city) does the gridfile I feed your program just need to contain
>the lat/lon of that one city?  Your code would then take the .nc file for
>all of North America and find the closest locations to my lat/lon and do
>the interpolation from there?  Is that how it would work?
>
>Thanks so much!
>Ethan
>
>On Wed, Jun 19, 2013 at 4:44 PM, Seth McGinnis <mcginnis at ucar.edu> wrote:
>
>> Hi Ethan,
>>
>> It regrids to whatever set of locations you've got in the gridfile.  It is
>> a
>> pretty slow program, and fewer output locations does go faster, so it's
>> good to
>> reduce it just to the specific locations you need if you can.
>>
>> Cheers,
>>
>> --Seth
>>
>>
>>
>> On Wed, 19 Jun 2013 15:42:36 -0400
>>  Ethan Bodnaruk <ebodnaruk at gmail.com> wrote:
>> >Hi Seth,
>> >
>> >Does your code regrid for the entire North America region or can it regrid
>> >for just a specific location (which would be much faster I would think)?
>> >
>> >Thanks!
>> >Ethan
>> >
>> >On Tue, Jun 18, 2013 at 7:30 PM, Seth McGinnis <mcginnis at ucar.edu> wrote:
>> >
>> >> Hi Darrin,
>> >>
>> >> I have been using kriging functions from the R library "fields" to
>> regrid
>> >> the
>> >> data.  (Fancy interpolation methods like kriging preserve the spatial
>> >> variability better than averaging-based methods like
>> >> inverse-distance-weighting
>> >> in mountainous and coastal regions.)
>> >>
>> >> You can find my code here:
>> http://www.narccap.ucar.edu/contrib/tools/R/
>> >>
>> >> (I've regridded a bunch of time-average data for temp and precip, and
>> have
>> >> plans to publish it as part of the NARCCAP archive, but it's not quite
>> >> polished
>> >> enough to make it available to the general public just yet.)
>> >>
>> >> Cheers,
>> >>
>> >> --Seth
>> >>
>> >>
>> >>
>> >> On Fri, 14 Jun 2013 11:42:56 -0600
>> >>  Darrin Sharp <dsharp at coas.oregonstate.edu> wrote:
>> >> >What's the easiest way to put all the NARCCAP results on a common grid
>> >> (i.e.
>> >> >with the same lat/lon points for all the models)? Is there a tool
>> >> specifically
>> >> >for this? Has somebody already done it?
>> >> >
>> >> >Trying to avoid writing this code myself - I know I'm not the first
>> >> person to
>> >> >need this.
>> >> >
>> >> >-Darrin
>> >> >
>> >> >======================================
>> >> >Darrin Sharp
>> >> >Faculty Research Assistant
>> >> >Oregon Climate Change Research Inst. (OCCRI)
>> >> >dsharp at coas.oregonstate.edu
>> >> >970-219-6713
>> >> >
>> >> >_______________________________________________
>> >> >narccap-discuss mailing list
>> >> >narccap-discuss at mailman.ucar.edu
>> >> >http://mailman.ucar.edu/mailman/listinfo/narccap-discuss
>> >>
>> >> _______________________________________________
>> >> narccap-discuss mailing list
>> >> narccap-discuss at mailman.ucar.edu
>> >> http://mailman.ucar.edu/mailman/listinfo/narccap-discuss
>> >>
>> >
>> >
>> >
>> >--
>> >Ethan Bodnaruk
>> >M.S. Nuclear Engineering
>> >Ph.D. Student, Ecological Engineering
>> >Author, *Beyond Religion: Science and Spirituality Aligned*
>> >www.ethanbodnaruk.com
>> >_______________________________________________
>> >narccap-discuss mailing list
>> >narccap-discuss at mailman.ucar.edu
>> >http://mailman.ucar.edu/mailman/listinfo/narccap-discuss
>>
>> _______________________________________________
>> narccap-discuss mailing list
>> narccap-discuss at mailman.ucar.edu
>> http://mailman.ucar.edu/mailman/listinfo/narccap-discuss
>>
>
>
>
>-- 
>Ethan Bodnaruk
>M.S. Nuclear Engineering
>Ph.D. Student, Ecological Engineering
>Author, *Beyond Religion: Science and Spirituality Aligned*
>www.ethanbodnaruk.com
>_______________________________________________
>narccap-discuss mailing list
>narccap-discuss at mailman.ucar.edu
>http://mailman.ucar.edu/mailman/listinfo/narccap-discuss



More information about the narccap-discuss mailing list