Namelist updates for the next cism main tag

[Katetc]

This PR is currently next up: ESCOMP/CISM#74

To support this in CISM-wrapper, we will need updates to the namelist_definition_cism.xml file and likely to test cases. We should update defaults so the tests are for useful configurations, and probably add some tests for new physics configurations.

The adjustments in this commit: 7823447
should probably be reversed if possible.

[whlipscomb]

@Katetc: Here's a list of namelist changes:

Changes in options

ocean_data_extrapolate: Reword the descriptions
integer :: ocean_data_extrapolate = 0
Old:

• [0] ocean data not extrapolated to shelf cavities
• [1] ocean data extrapolated to shelf cavities

New:

• [0] CISM will not extrapolate ocean data into cavities
• [1] CISM will extrapolate off-shelf ocean data into shelf cavities

smb_input_function: Add option 3

• [3] SMB computed with a positive-degree scheme

enable_acab_anomaly: Change name to enable_smb_anomaly

Changes in ho_options

which_ho_powerlaw_c: Add option 3

• [3] invert for basin-scale powerlaw_c

which_ho_coulomb_c: Change option 3, add option 4
Old:

• [3] coulomb_c = function of bed elevation

New:

• [3] invert for basin-scale coulomb_c_lo/hi
• [4] read basin-scale coulomb_c_lo/hi from external file

which_ho_coulomb_c_relax: Remove this option

elevation-based_coulomb_c: Add this option

• logical :: elevation_based_coulomb_c = .false.
• Flag that indicates whether coulomb_c depends on elevation
• (coulomb_c_hi for high bed, coulomb_c_lo for low bed, interpolated in between)

which_ho_bmlt_basin: Remove this option

which_ho_deltaT_ocn: Reword option 1, add option 3, move old (3) to (4)
Old:

• [0] deltaT_ocn = 0
• [1] invert for deltaT_ocn to match thickness target
• [2] read deltaT_ocn from external file
• [3] set deltaT_ocn to match dH/dt target

New:

• [0] deltaT_ocn = 0
• [1] invert for 2D deltaT_ocn to match thickness target
• [2] read deltaT_ocn from external file
• [3] invert for basin-scale deltaT_ocn
• [4] set deltaT_ocn to match dH/dt target

which_ho_effecpress: Reword (2) and (4), replace (3) with a new option
Old:

• [2] N is reduced where basal water is present, with a ramp function
• [3] N is reduced where there is a nonzero water flux at the bed
• [4] N is reduced where basal water is present, following Bueler/van Pelt

New:

• [2] N is reduced based on basal water thickness
• [3] N depends on cavity opening and closing
• [4] N is reduced based on basal water thickness, following Bueler/van Pelt

deltaT_ocn_extrapolate: Remove option

diva_slope_correction: Add option

• logical :: diva_slope_correction = .true.
• True if the DIVA solver includes a correction for steep slopes

which_ho_flotation_function: Change (4) to (3), remove the old (3). Change the default to (2).
Old:

• [3] Modified version of (2), without extrapolation to ice-free ocean
• [4] Like (3), but with stdev correction for topg

New:

• [3] Like (2), but with a correction for topg

New basal_hydro options

ho_flux_routing_scheme: Move to the basal_hydro group (otherwise unchanged)

• integer :: ho_flux_routing_scheme = 0

Two new basal_hydro_options:
cavity_open_slide

• integer :: cavity_open_slide = 0
• Treatment of opening by sliding in cavity-sheet equations
• [0] no opening by sliding
• [1] opening by sliding with fixed u_b
• [2] opening by sliding with dynamic u_b

cavity_open_melt

• integer :: cavity_open_melt = 0
• Treatment of opening by melting in cavity-sheet equations
• [0] no opening by melting
• [1] opening by melting based on bmlt_ground
• [2] opening by melting based on dissipation in cavities
• [3] opening by melting based on bmlt_ground plus dissipation

Parameter changes

Note: All of the following belong to the “parameters” group except those belonging to the new “basal_hydro” group.

glide_climate, new parameters:

• real(dp) :: degree_factor = 1000.d0 !> ablation rate as a function of temperature above tmlt (mm/yr w.e./deg C)
• real(dp) :: tmlt = -1.d0 !> spatially uniform temperature threshold for melting (deg C)
• real(dp) :: snow_threshold_min = 0.0d0 !> air temperature (deg C) below which all precip falls as snow
• real(dp) :: snow_threshold_max = 2.0d0 !> air temperature (deg C) above which all precip falls as rain

glide_climate, name changes, ‘acab_’ -> ‘smb_’:
Old

• real(dp) :: acab_factor = 1.0d0 !> adjustment factor for external acab field (unitless)
• real(dp) :: acab_anomaly_tstart = 0.0d0 !> time to start applying the anomaly (yr)
• real(dp) :: acab_anomaly_timescale = 0.0d0 !> number of years over which the acab/smb anomaly is phased in linearly

New

• real(dp) :: smb_factor = 1.0d0 !> multiplicative factor for external smb field (unitless)
• real(dp) :: smb_anomaly_tstart = 0.0d0 !> time to start applying the anomaly (yr)
• real(dp) :: smb_anomaly_timescale = 0.0d0 !> number of years over which the smb anomaly is phased in linearly

glide_inversion, new parameters:

• real(dp) :: babc_length_scale = 0.0d0, & !> diffusive length scale (m) for inversion
• real(dp) :: deltaT_ocn_length_scale = 0.0d0 !> diffusive length scale (m) for inversion
• real(dp) :: deltaT_basin_relax = 0.0d0, & !> relax the basin correction toward this value

glide_basal_physics, new parameters:

• real(dp) :: coulomb_c_const_hi = 0.50d0 !> constant coulomb_c value at high bed elevation, topg >= bed_hi
• real(dp) :: coulomb_c_const_lo = 0.10d0 !> constant coulomb_c value at low bed elevation, topg <= bed_lo
• real(dp) :: coulomb_c_bed_hi = 0.d0 !> bed elevation (m) above which coulomb_c = coulomb_c_hi
• real(dp) :: coulomb_c_bed_lo = -500.d0 !> bed elevation (m) below which coulomb_c = coulomb_c_lo

glide_ocean_data, new parameters:

• real(dp) :: thermal_forcing_basin_min = 0.0d0 !> min value of thermal_forcing_basin (deg C) applied to nonlocal and nonlocal-slope schemes

glide_basal_hydro (new section), new parameters:

• real(dp) :: bwat_threshold = 1.0d-3 !> scale over which N ramps down from overburden to a small value (m)
• real(dp) :: btemp_scale = 0.0d0 !> temperature scale (degC) for transition between thawed and frozen bed

glide_basal_hydro, parameters moved from other groups:

• real(dp) :: const_bwat = 10.d0 !> constant basal water depth (m)
• real(dp) :: effecpress_delta = 0.02d0 !> min value for effecpress N relative to overburden (unitless)
• real(dp) :: bpmp_threshold = 0.1d0 !> temperature range over which N ramps up from a small value to overburden (deg C)
• real(dp) :: bwat_till_max = 2.0d0 !> maximum water depth in till (m)
• real(dp) :: c_drainage = 1.0d-3 !> uniform drainage rate (m/yr)

Other changes

• Remove effecpress_bmlt_threshold
• Remove inversion_bmlt_basin_timescale
• Remove inversion_dbmlt_dtemp_scale
• Remove coulomb_c_relax_min, coulomb_c_relax_max, coulomb_c_bedmin, coulomb_c_bedmax
• Remove ocean_p_timescale
• remove_isthmuses: Change the default to “false” for Antarctica
• limit_marine_cliffs: Change the default to “false” for both ice sheets
• adjust_input_thickness: Change to “false” for Antarctica when we have new input files
• idiag: Change the gris4 default to 136
• jdiag: Change the gris4 default to 294
• max_slope: Change the default to 0.5 (if this is stable for Greenland)
• Change the name of “flow_factor” to “flow_factor_ground”

[Katetc]

A few ideas for new tests in CESM.

1. Add an AIS test with ocean_data_extrapolate=1
2. Phase in anomalies. Can we use this in CESM? Do we need files with forcing data to test these options?
3. Elevation_based_coulomb_c and corresponding hi and lo values. GRIS?
4. which_ho_effecpress options, currently only "0" is tested
5. Add a test with the Glide dycore in serial mode?
   Will add more later...

[whlipscomb]

@Katetc: Thanks, these are excellent ideas. A few thoughts on the specifics:

1. Definitely. This could be a CESM test where we extrapolate ocean data from the calving front into cavities.
2. We routinely phase in anomalies during the historical segment of glacier runs (following the baseline spin-up). Maybe we could have two glacier tests, one that tests spin-up with inversion, and one that tests anomaly forcing without inversion. Let's talk more about the best way to support glacier tests.
3. This could be part of a GrIS spin-up using the pseudo-plastic basal friction law.
4. I'd like to add a test with which_ho_bwat = 3 (steady-state flow routing) and which_ho_effecpress = 2 (reduce N based on active hydrology), since these two settings are linked.
5. Yes, I'd like to include a test for Glide. Do you mean a serial build, or an MPI build with 1 core? I haven't been maintaining the serial build (i.e., updating parallel_slap whenever I update parallel_mpi) and was thinking of dropping it.

Adding @gunterl to be part of the discussion.

To be continued...