Changes between Version 2 and Version 3 of 2020WP/KERNEL-01_Amy_Mike_newHPGschemes

Timestamp:

2020-06-04T13:57:03+02:00 (4 years ago)

Author:

ayoung

Comment:

Expanded description of proposed test case.

2020WP/KERNEL-01_Amy_Mike_newHPGschemes

@@ -59 +59 @@
 
 {{{#!box width=50em info
-These developments will be assessed using the idealised seamount configuration introduced by Beckmann & Haidvogel (1993).  This is a standard, widely used test case for evaluating the performance of hpg calculations.
+These developments will be assessed using the idealised seamount configuration introduced by Beckmann & Haidvogel (1993).  This is a standard, widely used test case for evaluating the performance of hpg calculations.  It consists of a Gaussian bump bathymetry initialised with the flow at rest, and an expontential vertical density perturbation profile.
 These developments should introduce no change to results unless the new hpg options are activated.
 If activated, the new schemes should result in a reduction of spurious currents in regions of sloping bathymetry when using terrain following coordinates (sco).
+The test configuration requires a bespoke set of MY_SRC files, a modification to the eosbn2.F90 module, and several namelist parameters.  This configuration currently exists as a UKMO branch from r4.0.2 (http://forge.ipsl.jussieu.fr/nemo/browser/NEMO/branches/UKMO/NEMO_4.0.2_SEAMOUNT) though does not yet run on multiple processors.
+There is no obvious single verification value.  Ideally, the result would be a reduction in error with respect to the standard sco formulation.  This may be defined in terms of a reduction in the maximum error (i.e. |umax_new| < |umax_sco|) or it may be a reduction in the time integrated value.  Depending on the scheme in question, there may also be results which should be identically zero.
 [[Include(wiki:Developers/DevProcess#tests)]]
 }}}