Module load_ncdf
Warning: we advise to use instead module local
This IGM module is designed to load spatial 2D raster data from a NetCDF file specified by the input_file parameter. The module converts all existing 2D fields into TensorFlow variables. At a minimum, the module is expected to import basal topography represented by the topg variable. Additionally, it completes the data, such as deriving basal topography from ice thickness and surface topography. Other fields present in the NetCDF file will also be converted to TensorFlow variables, allowing them to be accessed in the code via state.myvar. For example, providing the icemask variable can be useful in defining an accumulation area, which is beneficial for modeling individual glaciers and preventing overflow into neighboring catchments.
The module offers functions for resampling the data, where the coarsen parameter can be set to values like 2, 3, or 4 (with a default value of 1 indicating no coarsening). It also provides functionality for cropping the data by setting the crop parameter to True and specifying the desired bounds.
Additionally, by setting icemask_invert to True, an ice mask can be generated from an ESRI Shapefile specified by the icemask_shapefile parameter. This mask can identify areas that should contain glaciers or areas that should remain glacier-free, based on the icemask_include parameter.
The module also supports restarting an IGM run using a NetCDF file produced from a previous IGM run. To achieve this, provide the output NetCDF file from the previous run as input to IGM. The module will seek data corresponding to the starting time defined by processes.time.start and initialize the simulation at that time.
This module depends on netCDF4.
Contributors: G. Jouvet, A. Henz (icemask add-on).
Parameters
Default configuration file (load_ncdf.yaml):
load_ncdf:
input_file: input.nc
method_coarsen: skipping
coarsen: 1
crop: False
xmin: -1.0e+20
xmax: 1.0e+20
ymin: -1.0e+20
ymax: 1.0e+20
icemask_include: False
icemask_shapefile: icemask.shp
icemask_invert: False
Description of the parameters:
| Name | Description | Default value | Units |
|---|---|---|---|
input_file
|
NetCDF input data file. | input.nc | — |
method_coarsen
|
Method for coarsening the data from NetCDF file (skipping or cubic_spline). | skipping | — |
coarsen
|
Coarsen the data from NetCDF file by a certain (integer) number (2 would be twice coarser ignore data each 2 grid points). | 1 | — |
crop
|
Crop the data from NetCDF file with given top/down/left/right bounds. | False | — |
xmin
|
X left coordinate for cropping the NetCDF data. | -1e+20 | m |
xmax
|
X right coordinate for cropping the NetCDF data. | 1e+20 | m |
ymin
|
Y bottom coordinate for cropping the NetCDF data. | -1e+20 | m |
ymax
|
Y top coordinate for cropping the NetCDF data. | 1e+20 | m |
icemask_include
|
Include ice mask in the NetCDF file. | False | — |
icemask_shapefile
|
Shapefile for ice mask. | icemask.shp | — |
icemask_invert
|
Invert the ice mask. | False | — |