hydro_radial

The landice/hydro_radial test group implements variants of the radially symmetric hydrological test case from Bueler et al. (2015).

The domain is approximately rectangular with a radially symmetric ice sheet at its center. The mesh has 1-km horizontal resolution and 3 vertical layers. <<<Further description here>>>

<<<Exmaple image here>>>

The test group includes 4 test cases. All test cases are made up of 3 types of steps, setup_mesh, which defines the mesh and initial conditions for the model; run_model (given another name in many test cases to distinguish multiple forward runs), which performs time integration of the model; and visualize, which optionally plots the results of the test case (PNG files, plot windows, or both).

config options

All 4 test cases share the same set of default config options:

# config options for hydro_radial test cases
[hydro_radial]

# sizes (in cells) for the 1000m uniform mesh
nx = 50
ny = 58

# resolution (in m) for the 1000m uniform mesh
dc = 1000.0

# number of levels in the mesh
levels = 3


# config options related to visualization for hydro_radial test cases
[hydro_radial_viz]

# which time index to visualize
time_slice = -1

# whether to save image files
save_images = True

# whether to hide figures (typically when save_images = True)
hide_figs = True

spinup_test

landice/hydro_radial/spinup_test performs a 10,000-year spin-up run from “zero” (a very thin ice sheet) to quasi-steady state.

steady_state_drift_test

landice/hydro_radial/steady_state_drift_test performs a 1-month run, starting from the “exact” (that is, a nearly exact steady-state solution) as the initial condition.

decomposition_test

landice/hydro_radial/decomposition_test runs short (1-month) integrations of the model forward in time on 1 (1proc_run step) and then on 3 cores (3proc_run step) to make sure the resulting prognostic variables are bit-for-bit identical between the two runs.

restart_test

landice/hydro_radial/restart_test first run a 2-year integration of the model forward in time (full_run step). Then, a second step (restart_run) performs 2 1-year runs, where the second begins from a restart file saved by the first. Prognostic variables are compared between the “full” and “restart” runs at year 2 to make sure they are bit-for-bit identical.