Opened 10 years ago

Closed 10 years ago

Last modified 2 years ago

#613 closed Bug (fixed)

Problem with ORCA2 coordinates at NEMO vn2+

Reported by: idcatukmo Owned by: nemo
Priority: low Milestone: Unscheduled
Component: OCE Version: release-2
Severity: Keywords: ORCA2
Cc: nemo_st@…

Description

Dear all,

The central ORCA2 download page at

http://www.nemo-ocean.eu/Using-NEMO/Configurations/ORCA2_LIM

says

"CAUTION 2 versions are availbable:

  • ORCA2_LIM.tar could be used with all NEMO versions previous to the tag nemo_v2 i.e. downloaded before 12th november 2006
  • ORCA2_LIM_nemo_v2.tar is dedicated to NEMO versions tagged nemo_v2 and the following ones nemo_v2_xx"

etc.

Can someone tell me the reason for this change please? I ask because the u,v and f coordinates are now wrong in row 148 (=last physical row of ORCA2). The T-points are OK, I think.

Attached is a polar plot of some ORCA2 points near the north pole.
The frame on the left is from the old (pre-vn2) ORCA2 coordinates file;
the one on the right is from the new (post-vn2) ORCA2 coordinates file.

Consider point (135, 148). In this region of the grid i is increasing towards the top of the page and j is increasing towards the left of the page. This can be shown by printing out the coordinates of T(134:136, 148), and T(135, 147:149).

(Note that, because ORCA2 uses a t-pivot north fold, row 148 "doubles back on itself", so that, for instance, T-point (49, 148) coincides with T-point (135, 148). The picture shows this by also highlighting point (52, 148), where the i and j indices are increasing in the downward and rightward directions, respectively.)

Anyway, for the old coords file (left frame), U(135, 148) and V(135,
148) are in the right locations wrt T(135, 148), and the four f-points F(134:135, 147:148) surround T(135, 148), as required. Similarly for point (52, 148).

On the new grid (right frame), however, U(135, 148) and V(135, 148) are wrongly located wrt T(135, 148) (they match the orientations of the grid around (52, 148), where the indices are in the reverse direction, on account of the the doubling back of the grid.) The f-points are even worse: they lie on a line, instead of surrounding T(135,148). (Note that F(134, 147) = F(135, 148); in fact we should have F(134, 147) = F(49, 148).)

So: can anyone say why the ORCA2 coords file was changed at NEMO vn2? (I notice that the scale factors e1{u,v,f} and e2{u,v,f} are different too.) Is it safe to keep using the old one? It matters to us because we wish to use conservative interpolation, and to calculate the remapping coefficients, we need the corners of the cells, ie the f-points.

Thanks for any help,
Ian.

Commit History (0)

(No commits)

Attachments (2)

ORCA2_old_and_new_polar_coords.jpg (91.6 KB) - added by idcatukmo 10 years ago.
Picture of ORCA2 T,u,v,f points near N pole in pre-vn2 and post-vn2 NEMO
ORCA2_new_and_old.txt (895 bytes) - added by idcatukmo 10 years ago.
Coordinates of point (135,148) from both coordinate files.

Download all attachments as: .zip

Change History (8)

Changed 10 years ago by idcatukmo

Picture of ORCA2 T,u,v,f points near N pole in pre-vn2 and post-vn2 NEMO

Changed 10 years ago by idcatukmo

Coordinates of point (135,148) from both coordinate files.

comment:1 Changed 10 years ago by idcatukmo

  • Cc idcatukmo added
  • Priority changed from major to minor

comment:2 Changed 10 years ago by idcatukmo

  • Cc idcatukmo removed
  • Keywords ORCA2 coordinates added

comment:3 Changed 10 years ago by smasson

  • Resolution set to fixed
  • Status changed from new to closed

I checked your point and I 100% agree!

  • the two last lines of coordinates.nc given in ORCA2_LIM_nemo_v2.tar and ORCA2_LIM_nemo_v3.1.tar are wrong.
  • the original coordinates.nc that was in ORCA2_LIM.tar is the good one

⇒ ORCA2_LIM_nemo_v3.2.tar is containing (from the beginning) the coordinates.nc originally used in ORCA2_LIM.tar

Unfortunately, I was not able to trace back why and how the coordinates.nc was changed…
note: coordinates of other ORCA configurations are based on ORCA05 and should not be corrupted but we better check…

comment:4 Changed 4 years ago by nicolasmartin

  • Keywords coordinates removed

comment:5 Changed 2 years ago by nemo

  • Keywords Misc. added

comment:6 Changed 2 years ago by nemo

  • Keywords Misc. removed
Note: See TracTickets for help on using tickets.