Re: [newbie] 3D LCC: Cartesian, rectilinear, other?

From: Dennis Shea <shea_at_nyahnyahspammersnyahnyah>
Date: Mon Nov 26 2012 - 11:32:17 MST

I think what you are thinking of is

    float x(x) [ or double ]
    float y(y)
    float data(sigma,y,x)

whereas, I think the ESMF software, wants the lat and lon arrays

    float lat(y,x)
    float lon(y,x)

===
As I said below

> In NCL parlance, any grid where grid points are uniquely described
> by two dimensional arrays, eg, LAT(ny,ny) and LON(ny,nx) is called
> curvilinear.

Maybe, I should have written

In NCL parlance, any grid where grid points are uniquely described
by two dimensional arrays, latitude and longitude arrays: eg, LAT(ny,ny)
and LON(ny,nx) is called curvilinear.

I think the ESMF works on latitude and longitude arrays only.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Please ftp your global source file (netCDF or grib or hdf) and, if you
have it, a file containing the 2d latitude and longitude arrays
of the target grid. If you not have the latter, I think we can generate
a file.

ftp ftp.cgd.ucar.edu
anonymous
<use your email address for the password>
cd incoming
put ...
put ...
quit

Send email to me after you have successfully completed the transfer.

Note that for security reasons, you cannot list the contents of this
directory, and neither can I. I need to know the *exact* names of the
files in order to retrieve them.

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

On 11/25/12 11:00 PM, Tom Roche wrote:
>
> Tom Roche Wed, 21 Nov 2012 23:30:42 -0500
>>>>> I'm trying to select the appropriate method(s) for regridding gas
>>>>> concentrations from a global 3D grid (coordinates=longitude,
>>>>> latitude, sigma) to a regional 3D grid (coordinates=col, row,
>>>>> sigma) on a Lambert Conformal Conic (LCC) projection over North
>>>>> America with 12-km horizontal resolution.
>
>>>>> I suspect my difficulty is at least in part due to my uncertainty
>>>>> regarding definitions of some terms as used in the NCL docs. (It
>>>>> may also be due to my being totally confused :-) My global input
>>>>> grid is pretty clearly spherical and curvilinear (no?)
>
> Dennis Shea Sat, 24 Nov 2012 16:27:09 -0700
>> Logically rectilinear. Any grid that can be described by a
>> coordinate variable which, by definition, is a one dimensional
>> monotonically {in/de}creasing numeric array [eg: lat(lat) , lon(lon) ]
>> is a rectilinear array.
>
> OK, that seems reasonable. But ...
>
>>>>> I'm [even] less certain about the terms to use regarding my regional
>>>>> output grid: Does NCL consider a 3D LCC grid "Cartesian"? [...] If
>>>>> not Cartesian, I'm assuming a 3D LCC grid is at least considered
>>>>> rectilinear. If a 3D LCC grid is *not* considered rectilinear, I am
>>>>> very confused, and would appreciate an explanation.
>
> Dennis Shea Fri, 23 Nov 2012 08:47:30 -0700
>>>> [1] The LCC is curvilinear in NCL parlance.
>
> Tom Roche Sat, 24 Nov 2012 12:08:38 -0500
>>> I am very confused--this is the opposite of what I assumed.
>
>> In NCL parlance, any grid where grid points are uniquely described
>> by two dimensional arrays, eg, LAT(ny,ny) and LON(ny,nx) is called
>> curvilinear.
>
> But the LCC grids I'm working with are logically rectilinear, using your
> description above: each coordinate variable is one-dimensional and
> monotonic. Am I missing something?
>
>>> Is LCC at least "Cartesian"? That matters to me, because I'd like to
>>> do the 3D regridding with ESMF_regrid
>
>> Once the weight file is generated,
>> NCL's ESMF regridding operations will regrid any dimensional grid
>
> but the link you cite says
>
> http://www.ncl.ucar.edu/Document/Functions/ESMF/ESMF_regrid_with_weights.shtml
>>>>>> Using the provided weight file, regrids data from one lat/lon grid
>>>>>> to another.
>
> I need to 3D-regrid from lon-lat to LCC.
>
> TIA, Tom Roche <Tom_Roche@pobox.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 Mon Nov 26 11:32:27 2012

This archive was generated by hypermail 2.1.8 : Fri Dec 07 2012 - 13:30:06 MST