[gdal-dev] "reboxing," or 3D regridding

[Tom Roche]

http://lists.osgeo.org/pipermail/gdal-dev/2012-November/034680.html
>>> I'd appreciate advice regarding tools and methods for transforming
>>> values from an unprojected global 3D spatial grid to a projected 3D
>>> grid with different horizontal and vertical resolution, or pointers
>>> to other resources to consult.

http://lists.osgeo.org/pipermail/gdal-dev/2012-November/034689.html
>> I'd look for re-gridding code designed for modeling -- in particular
>> ESMF has a bunch of good stuff,

Yes! but ...

>> designed specifically for this sort of thing:

... alas, no. I'd post a link to their list, but esmf_support has a
private archive and I have not yet been granted subscriber status.
However my post to that list did receive the following reply:

Subject: [3587543] 3D regridding from spherical coordinates to Cartesian coordinates?
From: Robert Oehmke <robert.oehmke at noaa.gov>
Date: Fri, 16 Nov 2012 10:57:49 -0700
Cc: esmf_support <esmf_support at list.woc.noaa.gov>

> We currently only support regridding between [grids] with the same
> coordinate space (e.g. from Cartesian to Cartesian). If you convert
> [your input grid] to Cartesian then you could use the 3D regridding[;]
> is that a possibility for you?

Any suggestions, regarding how to conservatively "pseudo-project" my
global unprojected input into something Cartesian that I could use as
input to ESMFRegrid, would be appreciated. Details on the problem:

I'm working on N2O for CMAQ (an EPA-supported regional-scale model,
currently used for air quality applications, but growing in scope). This
is to be tested over North America using a domain used for AQMEII, an AQ
model intercomparison project run by US EPA and EU JRC

https://github.com/TomRoche/cornbeltN2O/wiki/AQMEII-North-American-domain#wiki-EPA

Pretty much all my N2O inputs are global, but all my other N2O inputs
are 2D, and numerous 2D APIs exist that for regridding unprojected to
projected. (Hence my puzzlement that this is so much more difficult in
3D, but my spatial consciousness is still immature.) For my boundary/
initial conditions, I'd like to use the output of a 3D spatial (i.e.,
horizontal+vertical, not horizontal+ temporal) global N2O model run by a
collaborator @ MIT CGCS.

I can crop the 3D model to have horizontal extents just larger than
AQMEII-NA (e.g., lon=-130.5° to -59°, lat=23° to 59°); i.e., I can
transform the unprojected global input to a cropped but still
unprojected regional input. But I'm wondering, how to get from the
cropped region to something that I could use as a Cartesian input to
ESMFRegrid? Apologies if the answer to this question is obvious; I'm a
masters student and far from spatially adept. And feel free to forward
me to some channel more suited for this request.

Your assistance is appreciated,
Tom Roche <Tom_Roche at pobox.com> <Roche.Tom at epa.gov>