[MITgcm-support] question about cube-sphere halos

[Alistair Adcroft]

Benny,

Yes - the values of any quantity in the corners of the halos are ambiguous
and so do not (should not) contain any data. Some parts of the algorithm
fill in the corners with data temporarily but only for calculation of local
terms (multi-dimensional advection is one example). The model should be
independent of whether the corners have zeros or not but in practice we put
zeros to avoid overflows.

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 Benny Cheng
Sent: Wednesday, February 11, 2004 8:04 PM
To: MITgcm-support at mitgcm.org
Subject: [MITgcm-support] question about cube-sphere halos


Is it natural to expect zeroes at the corners of the haloes, ie:

Here is a sample output of XC(-2:2,-2:2)  ( with overlap OLX = 3)

0.00000000000000D+000   0.00000000000000D+000   0.00000000000000D+000  
314.896705843670  314.943947509892
0.00000000000000D+000   0.00000000000000D+000   0.00000000000000D+000  
314.955059294431  315.000000000000
0.00000000000000D+000   0.00000000000000D+000   0.00000000000000D+000  
315.000000000000  315.044940705569
314.858553626787        314.922753304848        314.976968595699       
315.023031404301  315.077246695152
314.828462712491        314.900664842638        314.967693942090       
315.032306057910  315.099335157362

so, XC(-2:0,-2:0)  is a always block of zeroes  ? same question for other
corners.

- Benny


_______________________________________________
MITgcm-support mailing list
MITgcm-support at mitgcm.org
http://dev.mitgcm.org/mailman/listinfo/mitgcm-support