Re: I mean regridding the data to a rectiliniear lat / lon grid

As I mentioned earlier, I have no experience with Daymet tiles.
I think you will have to discover the best way to handle these.

What I meant was something like

   var(ntim,nlat,mlon) ; variable on Daymet tile netCDF file

   k = 4
   k2 = k/2
   var_avg = new ( (/ntim, nlat/k,mlon/k /), typeof(var), getFillValue(var))

   do nl=k2,nlat-3,k
     do ml=k2,mlon-3,k
        var_avg(:,nl,ml) = avg(var(:,nl-k2:nl+k2,ml-k2:ml+k2))
     end do
   end do

   LAT2D = lat2d(k2::k,k2::k) ; subset
   LON2D = lon2d(k2::k,k2::k)

Good luck
On 10/23/12 8:39 PM, Ping Yang wrote:
> Hi NCL,
>
> On Tue, Oct 23, 2012 at 4:12 PM, Dennis Shea <shea@ucar.edu
> <mailto:shea@ucar.edu>> wrote:
>
> Given the small area covered by the 2x2 tiles at 1km resolution.
> I think it would be just fine to compute a local 4kmx4km arithmetic
> average. The grid center would the mid point of the
> lat and lon used.
>
> What do you mean local 4kmx4km arithmetic average? Is this a regriding
> method?
>
> Looking forward to hearing from you.
>
> Regards,
>
> Ping
_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Wed Oct 24 09:24:32 2012