[MITgcm-support] question about cube-sphere halos

[Ichiro Fukumori]

We need halo information to interpolate to a point between the last
grid point of a tile and the first grid point of the neighboring tile.

An interpolation routine is needed for assimilation; e.g., interpolate
model fields to data points.  We also need to do a
user-defined-coarse-grid to model grid mapping for the Kalman filter.
Assigning nearest grid point is inadequate for this purpose.

For interpolation purposes, one possible hack at the corner of the
cube is to assign the neighboring grids either on the tile above or
the tile to the side.  We'll assign slightly shifted locations for
these halo points (say 100m) so that the interpolation routine doesn't
hiccup with duplicate node locations.  What do you think?

-- Ichiro

At 12:39 PM 2/13/2004 -0800, Benny Cheng wrote:
>Alistair. Thanks for the info. It's all I need for now.
>It's a long story, but basically, we need the halos to interpolate
>uniquely onto each tile.
>
>
>On Fri, 2004-02-13 at 12:21, Alistair Adcroft wrote:
> > I guess we need to know what interpolation is being done on and for - 
> why do
> > you have output within the corners and why are interpolating in the halo
> > regions when the data is on the next tile?
> >
> > A.
> > --
> > Dr Alistair Adcroft            http://www.mit.edu/~adcroft
> > MIT Climate Modeling Initiative        tel: (617) 253-5938
> > EAPS 54-1523,  77 Massachusetts Ave,  Cambridge,  MA,  USA
> >
> > -----Original Message-----
> > From: mitgcm-support-bounces at mitgcm.org
> > [mailto:mitgcm-support-bounces at mitgcm.org] On Behalf Of Alistair Adcroft
> > Sent: Friday, February 13, 2004 3:13 PM
> > To: mitgcm-support at mitgcm.org
> > Subject: RE: [MITgcm-support] question about cube-sphere halos
> >
> >
> > No - there is no correct values of XC or YC in the corner regions - what
> > value should XC(0,0,bi,bj) have? There are two possible different values.
> >
> > I didn't know about the CPP flag - I'll see what Chris is trying to do 
> there
> > but whatever it is it can't be done on the cube ...
> >
> > A.
> > --
> > Dr Alistair Adcroft            http://www.mit.edu/~adcroft
> > MIT Climate Modeling Initiative        tel: (617) 253-5938
> > EAPS 54-1523,  77 Massachusetts Ave,  Cambridge,  MA,  USA
> >
> > -----Original Message-----
> > From: mitgcm-support-bounces at mitgcm.org
> > [mailto:mitgcm-support-bounces at mitgcm.org] On Behalf Of Dimitris Menemenlis
> > Sent: Friday, February 13, 2004 3:05 PM
> > To: mitgcm
> > Subject: RE: [MITgcm-support] question about cube-sphere halos
> >
> >
> > Alistair, I notice a flag called EXCH_UPDATE_CORNERS in the exch2 routines.
> > Benny is only looking to get the values of XC, YC, XG, YG, for 
> interpolation
> > purposes (as his informative message clearly stated ;-).  These are not
> > ambiguous and setting them to correct values should not affect rest of code
> > in any way, right?
> >
> > D.
>
>_______________________________________________
>MITgcm-support mailing list
>MITgcm-support at mitgcm.org
>http://dev.mitgcm.org/mailman/listinfo/mitgcm-support

===============================================================
Ichiro Fukumori               | Jet Propulsion Laboratory
e-mail: fukumori at jpl.nasa.gov | Mail Stop 300-323
phone: +1 (818) 354-6965      | 4800 Oak Grove Drive
fax: +1 (818) 393-6720        | Pasadena, California 91109, U.S.A.
===============================================================