[MITgcm-devel] changes to layers output

[Jean-Michel Campin]

Hi Ryan,

I am not convinced this change is needed:
1) it does not seems too bad for me to have to multiply by DYG or DXG
 to get the volume tranport: it's already what I do when computing
 eulerian tranport from UVELMASS,VVEMASS or bolus transport from
 GM_PsiX and GM_PsiY. 
 On the other hand, if I want to compute, from the layer pkg diagnostics, 
  u'h' = < uh > - <u> <h>
 I would now have to divide by DYG (or DXG) ?
2) Changing the definition of a diagnostics is potentially a cause
 of problems. So, if we can avoid this, it would be better.
 Alternatively, we can change also the diagnostics names, but not so sure
 it's what everybody wants. And finally, there is also the question of
 what to do with output from pkg/timeave: keep them unchanged and loose
 the link with corresponding diagnostics or also change them (with no way
 to indicate that the content of the file has changed).

Cheers,
Jean-Michel

On Tue, Oct 28, 2014 at 11:01:30AM -0400, Ryan Abernathey wrote:
> Hi Folks,
> 
> This is a message for those of you who use the layers package.
> 
> I would like to change the output of layers to make it easier to compute
> volume transports. Currently we output the thickness-weighted transport as,
> e.g.
> 223 |LaVH1TH | 42 |   222 |VV      MX|m.m/s           |Layer Integrated
> merid. Transport (VH, m^2/s)
> I would like to modify UH and VH by multiplying them with DYG and DXG,
> giving units of m^3 / s. Currently this has to be done as a post-processing
> step, but I think it would be better to do it online.
> 
> If you are currently using these fields in your diagnostics, this change
> would probably break your analysis routines.
> 
> Please let me know if you object to the idea.
> 
> -Ryan

> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel