User Tools

Site Tools


noresm:testlist

Test list for NorESM

This page contains a list of tests which need to pass before any changes can be ported to the trunk. By definition, the trunk always passes these tests. Any person who finds that the trunk does not pass these tests should immediately send an email to noresm-ncc(at)met.no

Any development version which does not pass the tests can only exist on branches.

Restart test

  1. Start a 2 month branch run, create monthly average output
  2. Start the same model run, but define it as two single months with restart in between.
  3. Verify that the last monthly average is equal in both runs.

Bit-identical meteorology for "technical only" code changes

This test is applicable for code which is not supposed to change the model physics (e.g. writing out extra diagnostics, cleaning up code without changing the functionality..)

  1. Run the code without the new changes
  2. Run the code with the new changes
  3. Verify that temperature is equal in both runs (ncdiff file1.nc file2.nc file3.nc) should give zero values in file3.nc

Physical tests (early development)

These tests are applicable for early development of the model. At later stages, closer to important deliveries, other (and stricter) tests may apply.

Compare your new result with a result from the previous version of the trunk code.

If any of the global average of the following variables change with more than a limit X, then you must first pass through a discussion with the other developers. Send an e-mail to noresm-ncc(at)met.no before merging your changes to trunk:

  1. Aerosol optical depth (AOD_VIS, X=10%)
  2. Cloud droplet number concentration (CDNC, X=10%)
  3. Temperature (X = 2K)
  4. Liquid water path (LWP, X=10%)
  5. Total aerosol number concentration (N_AER, X=10%)
  6. Total precipitation (PRECT, X=10%)
noresm/testlist.txt · Last modified: 2013-10-31 13:57:50 by alfg