Implements an integration test based on the 25 km Arctic Ocean simulation that has been used since the June 2023 metting in Bergen. A script generates the restart and forcing files using python (since ncgen refused to parse the output of ncdump). The model runs and the output restart file is examined to ensure that a basic physical state exists (see below). The script cleans up the generated files from the test directory and returns the result of the test script to the CI workflow.
Test Description
The initial conditions are derived from TOPAZ4 data for 2010-01-01. The forcing data is derived from both ERA5 and TOPAZ4 interpolated to a Arctic Ocean grid, with one sample at 2010-01-01T00:00:00Z. This is duplicated to avoid a problem with forcing files containing only one sample. The model runs with BBM dynamics and a 300 s timestep for an hour (3600 s) of model time. The final restart file is then examined using a python script to ensure that a physical amount of ice exists at certain selected points, i.e. a thickness between 0 and 10 m, a concentration between 0 and 1 and reasonable values for the first DG component.
Integration test
Fixes #713
Change Description
Implements an integration test based on the 25 km Arctic Ocean simulation that has been used since the June 2023 metting in Bergen. A script generates the restart and forcing files using python (since
ncgen
refused to parse the output ofncdump
). The model runs and the output restart file is examined to ensure that a basic physical state exists (see below). The script cleans up the generated files from the test directory and returns the result of the test script to the CI workflow.Test Description
The initial conditions are derived from TOPAZ4 data for 2010-01-01. The forcing data is derived from both ERA5 and TOPAZ4 interpolated to a Arctic Ocean grid, with one sample at 2010-01-01T00:00:00Z. This is duplicated to avoid a problem with forcing files containing only one sample. The model runs with BBM dynamics and a 300 s timestep for an hour (3600 s) of model time. The final restart file is then examined using a python script to ensure that a physical amount of ice exists at certain selected points, i.e. a thickness between 0 and 10 m, a concentration between 0 and 1 and reasonable values for the first DG component.