1.3 GLIDE

GLIDE is the actual ice sheet model. GLIDE comprises three procedures which initialise the model, perform a single time step and finalise the model. The GLIDE configuration file is described in Section 1.3.1. The GLIDE API is described in Appendix B.2. The simple example driver explains how to write a simple climate driver for GLIDE. Download the example from the GLIMMER website or from CVS:

1.3.1 Configuration

The format of the configuration files is similar to Windows .ini files and contains sections. Each section contains key, values pairs.

• Empty lines, or lines starting with a #, ; or ! are ignored.
• A new section starts with the the section name enclose with square brackets, e.g. [grid].
• Keys are separated from their associated values by a = or :.

Sections and keys are case sensitive and may contain white space. However, the configuration parser is very simple and thus the number of spaces within a key or section name also matters. Sensible defaults are used when a specific key is not found.

[grid]
Define model grid. Maybe we should make this optional and read grid specifications from input netCDF file (if present). Certainly, the input netCDF files should be checked (but presently are not) if grid specifications are compatible.
ewn	(integer) number of nodes in x–direction
nsn	(integer) number of nodes in y–direction
upn	(integer) number of nodes in z–direction
dew	(real) node spacing in x–direction
dns	(real) node spacing in y–direction
sigma_file	(string) Name of file containing σ coordinates. The σ coordinates are calculated if no file name is given using the formula We should probably allow n to be a run–time parameter.
[time]
Configure time steps, etc. Update intervals should probably become absolute values rather than related to the main time step when we introduce variable time steps.
tstart	(real) Start time of the model in years
tend	(real) End time of the model in years
dt	(real) size of time step in years
ntem	(real) time step multiplier setting the ice temperature update interval
nvel	(real) time step multiplier setting the velocity update interval
[options]
Parameters set in this section determine how various components of the ice sheet model are treated. Defaults are indicated in bold.
ioparams	(string) name of file containing netCDF I/O configuration. The main configuration file is searched for I/O related sections if no file name is given (default).
temperature	0 isothermal 1 full
flow_law	0 Patterson and Budd 1 Patterson and Budd (temp=-10degC) 2 const 10-16a-1Pa-n
basal_water	0 local water balance 1 local water balance + const flux 2 none
marine_margin	0 ignore marine margin 1 Set thickness to zero if floating 2 Set thickness to zero if relaxed bedrock is more than certain water depth 3 Lose fraction of ice when edge cell
slip_coeff	0 zero 1 set to a non–zero constant everywhere 2 set constant where the ice base is melting 0 ∝ basal water
evolution	0 pseudo-diffusion 1 ADI scheme 2 diffusion
vertical_integration	0 standard 1 obey upper BC
topo_is_relaxed	0 relaxed topography is read from a separate variable 1 first time slice of input topography is assumed to be relaxed 2 first time slice of input topography is assumed to be in isostatic equilibrium with ice thickness.
periodic_ew	0 switched off 1 periodic lateral EW boundary conditions (i.e. run model on torus)
hotstart	Hotstart the model if set to 1. This option only affects the way the initial temperature and flow factor distribution is calculated.
[parameters]
Set various parameters.
log_level	(integer) set to a value between 0, no messages, and 6, all messages are displayed to stdout. By default messages are only logged to file.
ice_limit	(real) below this limit ice is only accumulated, ice dynamics are switched on once the ice thickness is above this value.
marine_limit	(real) all ice is assumed lost once water depths reach this value. Note, water depth is negative.
calving_fraction	(real) fraction of ice lost due to calving.
geothermal	(real) constant geothermal heat flux.
flow_factor	(real) the flow law is enhanced with this factor
hydro_time	(real) basal hydrology time constant
isos_time	(real) isostasy time constant
basal_tract_const	constant basal traction parameter. You can load a nc file with a variable called soft if you want a specially variying bed softness parameter. You can use construct_field.py (see Chapter 2.3.2) to produce a suitable input file.
basal_tract	(real(5)) basal traction factors. Basal traction is set to B = tanh(W) where the parameters (1) width of the tanh curve (2) W at midpoint of tanh curve [m] (3) B minimum [ma-1Pa-1] (4) B maximum [ma-1Pa-1] (5) multiplier for marine sediments
[isostasy]
Isostatic adjustment is only enabled if this section is present in the configuration file. The options described control isostasy model.
lithosphere	0 local lithosphere, equilibrium bedrock depression is found using Archimedes’ principle 1 elastic lithosphere, flexural rigidity is taken into account
asthenosphere	0 fluid mantle, isostatic adjustment happens instantaneously 1 relaxing mantle, mantle is approximated by a half-space
relaxed_tau	characteristic time constant of relaxing mantle (default: 4000.a)
update	lithosphere update period (default: 500.a)
[projection]
Specify map projection. The reader is referred to Snyder J.P. (1987) Map Projections - a working manual. USGS Professional Paper 1395.
type	This is a string that specifies the projection type (LAEA, AEA, LCC or STERE).
centre_longitude	Central longitude in degrees east
centre_latitude	Central latitude in degrees north
false_easting	False easting in meters
false_northing	False northing in meters
standard_parallel	Location of standard parallel(s) in degrees north. Up to two standard parallels may be specified (depending on the projection).
scale_factor	non-dimensional. Only relevant for the Stereographic projection.
[elastic lithosphere]
Set up parameters of the elastic lithosphere.
flexural_rigidity	flexural rigidity of the lithosphere (default: 0.24e25)
[GTHF]
Switch on lithospheric temperature and geothermal heat calculation.
num_dim	can be either 1 for 1D calculations or 3 for 3D calculations.
nlayer	number of vertical layers (default: 20).
surft	initial surface temperature (default 2∘C).
rock_base	depth below sea-level at which geothermal heat gradient is applied (default: -5000m).
numt	number time steps for spinning up GTHF calculations (default: 0).
rho	The density of lithosphere (default: 3300kg m-3).
shc	specific heat capcity of lithosphere (default: 1000J kg-1 K-1).
con	thermal conductivity of lithosphere (3.3 W m-1 K-1).

netCDF I/O can be configured in the main configuration file or in a separate file (see ioparams in the [options] section). Any number of input and output files can be specified. Input files are processed in the same order hey occur in the configuration file, thus potentially overwriting priviously loaded fields.