[MITgcm-devel] RE: Higher order dissip on the cube

[Alistair Adcroft]

So we should check the loop ranges in the new dissipation routines -
again...

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-devel-bounces at mitgcm.org
[mailto:mitgcm-devel-bounces at mitgcm.org] On Behalf Of Chris Hill
Sent: Sunday, February 08, 2004 10:36 AM
To: MITgcm-devel at mitgcm.org
Subject: RE: [MITgcm-devel] RE: Higher order dissip on the cube


J-M etc...,

 I took a quick look at the array syntax that got zapped and
 I made some progress but still have at least one mystery.

 I restored the unowned point loop that got zapped when the array syntax was
removed. This is in the attached exch2_uv_xyz_rl.F and _xy_ . To get
identical numbers to exch1 with bh I currently 
  1. zero out the null regions for u and v at the end of exch2 _xyz_ 
  2. avoid using exch2 _xy_, since when I do the zero'ing in _xy_ I get
NAN's on the second time step and if I leave out the zero'ing I don't get
the exact same numbers.

 Not sure exactly what that means yet. I think it points to some loop bounds
assumption thing possibly like the GM thing we hunted down before. I'm just
trying to hunt it down.

Chris

P.S. I'm testing with global_ocean.cs

> -----Original Message-----
> From: mitgcm-devel-bounces at mitgcm.org 
> [mailto:mitgcm-devel-bounces at mitgcm.org] On Behalf Of Jean-Michel 
> Campin
> Sent: Saturday, February 07, 2004 3:47 PM
> To: adcroft at mit.edu
> Cc: MITgcm-devel at mitgcm.org
> Subject: [MITgcm-devel] RE: Higher order dissip on the cube
> 
> Hi Alistair,
> 
> > I don't know where the bug is
> 
> I did nice plots to show where the differences are and sent 2 emails 
> to MITgcm-devel arround Jan 25. I just put here below the response of 
> Chris + original message.
> 
> Jean-Michel
> --------------------------------------------------------------
> -----------------
> Date: Mon, 26 Jan 2004 09:09:20 -0500
> From: Chris Hill <cnh at mit.edu>
> Reply-To: MITgcm-devel at mitgcm.org
> To: MITgcm-devel at mitgcm.org
> Subject: RE: [MITgcm-devel] Re: Problem with Exch2
> 
> JM,
> 
>  The numbers should still be valid for uvel(snx+1,-1) and they should 
> be the same on one or n processors [but they will not be valid for 
> vvel(snx+1,-1)]. Can you talk with Andy so he can see what you mean 
> and can see if something broke in check in.
> 
> Chris
> 
> > -----Original Message-----
> > From: mitgcm-devel-bounces at mitgcm.org 
> > [mailto:mitgcm-devel-bounces at mitgcm.org] On Behalf Of Jean-Michel 
> > Campin
> > Sent: Sunday, January 25, 2004 7:06 PM
> > To: MITgcm-devel at mitgcm.org
> > Subject: [MITgcm-devel] Re: Problem with Exch2
> > 
> > 
> > > the exhc2 does not give the same uVel(sNx,-1) as the old exch_uv.
> > 
> > The problem is in (sNx+1,-1) and also in (sNx+1,sNy+2) but
> not not in
> > uVel(sNx,-1).
> > Sorry for this.
> > 
> > Jean-Michel
> > _______________________________________________
> > MITgcm-devel mailing list
> > MITgcm-devel at mitgcm.org
> > http://dev.mitgcm.org/mailman/listinfo/mitgcm-devel
> > 
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://dev.mitgcm.org/mailman/listinfo/mitgcm-devel
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org 
> http://dev.mitgcm.org/mailman/listinfo/mitgcm-devel
>