[MITgcm-support] r*FacC crash

[Jean-Michel Campin]

Hi Dimitri & Baylor,

On Thu, May 18, 2006 at 09:57:42AM -0700, Dimitris Menemenlis wrote:
> Salut Jean-Michel, could you explain in a few sentences what an r*FacC 
> crash means and what can be the possible causes.  Thanks, Dimitris
> 
> >Hi Dimitris,
> >  I'm not familiar with the r*FacC crash.  What does it mean?
> >   -Baylor
> 
> >WARNING: r*FacC < hFacInf at: i,j,bi,bj,Thid,Iter=   5  39   1   1   1    
> >502399
> >rStarFac,H,eta = -0.707575  2.000000E+01 -3.415149E+01
> >STOP in CALC_R_STAR : too SMALL rStarFacC !

This means that the total column thickness is too small:
 H+eta < H * hFacInf
with: H = fixed depth (>0), from the bathymetry file
      hFacInf = parameter (in namelist PARM01, file "data")

And when this happens, it's safer to stop since the rescaled coordinate
may result in a "too small hFacC" : hFac(it=n) = hFac_initial * rStarFac
 with rStarFac=(H+eta)/H

In your case, H+eta is negative (no water left in this column i=5,j=39), 
and it's time to stop the model.

Jean-Michel