[MITgcm-devel] Re: odds and ends

[Jean-Michel Campin]

Hi Daniel,

Others might be interested, so I cc to mitgcm-devel list.

On Thu, Nov 17, 2005 at 04:46:42PM -0500, Daniel Enderton wrote:
> Hey Jean -Michel!
> 
> Hope your having a good time in Belgium.
> 
> If you have time for a couple questions...
> 
> (1)  I am trying to get all your key cube sphere functions a bit 
> cleaned up and user friendly to put in the cs_grid directory.  I have 
> banged out the heat transport script and the overturning script.  I 
> was working on the horizontal stream function script (calc_psiH_CS.m) 
> and had a couple questions.  Why is the output of the function of 
> length 6146 = 192*32+2 rather than 6144 = 192*32.  What are the two 
> extra points for?  Are there any particular two points that I can 
> strip off and plot using merccube?

Psi is located at grid-cell corner (= vorticity point) and the CS grid 
has 2 more corner points than cell-center points. To make a plot,
I usually interpolate to lat-lon (just need the list of longitude
& latitude corresponding to the Psi values) and then draw contour lines.

> (2) Is there anyway that I can generate u*phi and v*phi to get the 
> heat transport in the atmosphere associates with potential energy? 

I think Andrea checked-in 2 new diagnostics (UVELPHI & VVELPHI) 
two days ago.

> Further, since we are using zstar in the atmosphere we need hu and hv 
> to generate the overturning stream function.  Would this not mean we 
> should use hu*t and hv*t for the sensible heat transport.  I looked, 
> but saw not huttave (or similar) field in the output or in the 
> diagnostics package.

With timeave pkg, the hfac is already accounted for in the
output files UTtave & VTtave. With the diagnostics pkg, both
diagnostics are available (with hfac: UTHMASS,VTHMASS
 and without: UVELTH,VVELTH)

Cheers,
Jean-Michel