[MITgcm-devel] An overdue revision of some diagnostic variable names and their descriptions

[Fenty, Ian G (329C)]

Jean-Michel,

I think that changing some variable descriptions would be enough to overcome most of the confusion.
It was pointed out to me that the velocities in UVELMASS/VVELMASS are “essentially are mass weighted given that we are using the boussinesq approximation”.  I can’t argue with that....

Someone (probably Gael) used different descriptions for some of the variables in the ECCO Central Production README file(ftp://ecco.jpl.nasa.gov/Version4/Release3/nctiles_monthly/README).

Here is a somewhat edited snippet of some very well described fields:

Vm_dPHdy                V momentum tendency from Hydrostatic Pressure grad (m/s^2) along the model’s y-direction but may not be east.
VVEL                           V component of velocity (m/s) along the model's y-direction but may not be north. NVEL is the north velocity.
VVELMASS                 V geometry-weighted component of velocity (m/s) along the model's y-direction but may not be north.  VVELMASS is VVEL times hFacS to account for

Other fields are given a more succinct but still accurate description.

  SIuice                            SEAICE x-dir. ice velocity (m/s)
  SIvice                            SEAICE y-dir. ice velocity (m/s)

And still others mention “west” and “south” faces, faces that we all know refer to the “left” and “bottom” faces of the Arakawa-C tracer cell but that may not be obvious to a new user:

  ADVxHEFF              U Comp. Advective Flux of eff ice thickness (m.m^2/s) at the western face. Positive to increase eff ice thickness.
  ADVx_SLT               U Comp. Advective Flux of Salinity (psu.m^3/s) at the western face. Positive to increase salt.


If space in the description field is not an issue, then my vote is for being as descriptive as possible (i.e., following the example of the first snippet).  If space is an issue then if something like the examples below would fit then it would be enough,

VVELMASS                 V geometry-weighted velocity (m/s) in y-dir. VVELMASS = VVEL x hFacS to account for varying cell height (hFacS between 0 and 1).

As for the ‘west’ and ‘south’ faces mentioned in the description, perhaps references to the C-grid would suffice, such as

ADVxHEFF              U-component advective flux of average ice thickness (m.m^2/s) through the western face of the C-grid tracer cell.
ADVyHEFF              V-component advective flux of average ice thickness (m.m^2/s) through the southern face of the C-grid tracer cell.

But there may be even clearer methods.

What do you think?

Ian






On 1/16/18, 3:42 PM, "MITgcm-devel on behalf of Jean-Michel Campin" <mitgcm-devel-bounces at mitgcm.org<mailto:mitgcm-devel-bounces at mitgcm.org> on behalf of jmc at mit.edu<mailto:jmc at mit.edu>> wrote:

Hi Ian,

Any update that can improve/avoid confusion is welcome !

Changing some diagnostic descriptions is easy, we just need to agree on what
to put instead (and may be some consistency across diagnostic descriptions).

If you want also to change diagnostic names (e.g., "UVELMASS" --> ? ), it's
a little bit more tricky since many users are now used to specify these (bad)
8-characters diagnostic names in their data.diagnostics, so the transition
might be a little painful.

And just to bring an historical perspective, "UVELMASS", "VVELMASS" names
and description were added early on (Mars 11, 2004); from an atmospheric
point of view, using pressure coordinate, the "mass weighted" makes more
sense (hFac capture the time-dependence of the grid-box mass content).

Cheers,
Jean-Michel

On Tue, Jan 16, 2018 at 07:02:30PM +0000, Fenty, Ian G (329C) wrote:
Hi MITgcm devel,
In an effort to make our output more comprehensible to outsiders I think we should consider updating the names of some of our diagnostic output descriptions.
First, terms such as UVELMASS and UTHMASS are not ???Zonal components of velocity???, they are the component of velocity in the model grid???s ???x??? direction.  In the LLC tiles, 6 of the 13 tiles (or 2 of 5 faces) are rotated by 90 degrees so that flow in the ???+x??? direction is actually negative quasi-meridional.  Moreover, with the Arctic Cap tile 7 (face 3) flow in the positive ???x??? direction is eastward along its bottom edge (first row) and westward along its upper edge (last row).   A similar story applies for flow in the +y direction.  So, can we change these names to reflect the fact that ???u??? and ???v??? are defined not by the compass but by the local orientation of the curvilinear grid?
Second, the name ???MASS??? and the description ???Mass-Weighted??? is also somewhat misleading because these terms are not weighted by mass at all, but by hFac.  Perhaps a more descriptive term would be ???Geometry-Weighted??? or some other name that reflects the possible non-unity value of hFac?
This topic has come up many times since we???ve adopted the LLC grids for the ECCO Central Production and other projects.  I???d like to see this sorted out to reduce the confusion that seems to hit everybody encountering these variables for the first time.
   *   Ian
45 |UVELMASS| 50 |    46 |UUr     MR|m/s             |Zonal Mass-Weighted Comp of Velocity (m/s)
46 |VVELMASS| 50 |    45 |VVr     MR|m/s             |Meridional Mass-Weighted Comp of Velocity (m/s)
50 |UTHMASS | 50 |    51 |UUr     MR|degC.m/s        |Zonal Mass-Weight Transp of Pot Temp
51 |VTHMASS | 50 |    50 |VVr     MR|degC.m/s        |Meridional Mass-Weight Transp of Pot Temp

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

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

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.mitgcm.org/pipermail/mitgcm-devel/attachments/20180123/9b053dbf/attachment-0001.html>