Model Input Files
The FRICOSIPY model requires three input Network Common Data Format (NetCDF) files in order to run a simulation: Static, Meteo and Illumination. These replace the large 3-dimensional input dataset \((x,y,t)\) of the original COSIPY model in order to lessen the required computational resources and enable the execution of a simulation with a high spatio-temporal resolution. In FRICOSIPY, the meteorological data is instead interpolated across the spatial grid according to topographic variation during the main simulation.
\((1)\) Static File
The model static input file contains topographic information that varies across the spatial domain
\((x,y)\) and requires the following variables:
- NORTHING – Northing [m]
- EASTING – Easting [m]
- LATITUDE – Latitude (WGS84) [decimal]
- LONGITUDE – Longitude (WGS84) [decimal]
- ELEVATION – Elevation [m a.s.l.]
- ASPECT – Terrain aspect [°]
- SLOPE – Terrain slope [°]
- MASK – Glacier mask boolean [0 or 1]
Optional Variables:
The following optional variables can also be used:
- BASAL – Basal / geothermal heat flux [Wm\(^{-2}\)]
(enables a spatially-variable basal/geothermal heat flux and overrides the 'basal_heat_flux' parameter) - THICKNESS – Glacier thickness [m]
(enables a spatially-variable initial glacier height and overrides the 'initial_height' parameters) - PRECIPITATION_CLIMATIOLOGY – Precipitation climatology [m yr\(^{-1}\)]
(annual precipitation climatology for the three-phase anomaly model)
The static file can either be directly created point grid data in .CSV format or from a Digital Elevation Model (DEM) in .GeoTIFF format. The latter approach is easier and does not require the use of Geographic Information System (GIS) software but it is less versatile and does not enable the inclusion of any optional static variables.
\((i)\) Conversion from CSV to NetCDF
An exemplar static CSV would have the following format:
| NORTHING | EASTING | LATITUDE | LONGITUDE | ELEVATION | ASPECT | SLOPE | MASK |
|---|---|---|---|---|---|---|---|
| 1086500 | 2633800 | 45.92925 | 7.874360 | 4457.10 | 240.24 | 2.35 | 1 |
| 1086500 | 2633850 | 45.93039 | 7.874485 | 4456.72 | 210.12 | 4.13 | 1 |
| ⋮ | ⋮ | ⋮ | ⋮ | ⋮ | ⋮ | ⋮ | ⋮ |
| 1086800 | 2633800 | 45.93208 | 7.874285 | 4388.20 | 30.26 | 20.12 | 1 |
| 1086800 | 2633850 | 45.93195 | 7.875026 | 4358.10 | 36.78 | 23.21 | 1 |
Attention
FRICOSIPY requires a standard rectilinear grid. However, the northing and easting values can simply be substituted for a locally referenced grid since they simply form the spatial structure of the model and do not influence the physical processes.
Place the input CSV file in the data/static/CSV/ directory and then the create_static_netcdf_from_CSV.py utility program can then convert it into NetCDF format.
The program is launched, by navigating to the utilities/create_STATIC/ directory on the command line and executing the program create_static_netcdf_from_CSV.py with the following arguments:
- -c <static_csv>.csv – input static CSV file name
- -s <static_netcdf>.nc – output static NetCDF file name
- -p 'EPSG:<XXXX>' – (optional) spatial projection / co-ordinate reference system
cd utilities/create_STATIC/
\vspace{0.5cm}
python3 create_static_netcdf_from_CSV.py -c <static_csv>.csv -s <static_netcdf>.nc
\((ii)\) Conversion from GeoTIFF to NetCDF
Alternatively, place the input GeoTIFF file in the data/static/GeoTIFF/ directory and then the create_static_netcdf_from_GeoTIFF.py utility program can then convert it into NetCDF format. A shapefile (.shp) is also required to deliniate the boundary of the glacier (the glacier mask) which should be placed in the data/static/SHP/ directory.
The program is launched, by navigating to the utilities/create_STATIC/ directory on the command line and executing the program create_static_netcdf_from_GeoTIFF.py with the following arguments:
- -g <static_dem>.tif – input GeoTIFF file name
- -s <static_netcdf>.nc – output static NetCDF file name
- -m <mask>.shp – input glacier mask shapefile name
- -r <value> – (optional) resampled static file spatial resolution
cd utilities/create_STATIC/
python3 create_static_netcdf_from_GeoTIFF.py -g <static_dem>.tif -s <static_netcdf>.nc -m <glacier_mask>.shp
In Switzerland, high resolution topographic data is available through the SwissAlti3D Digital Elevation Model (DEM) product of the Federal Office of Topography (Swiss Topo).
\((2)\) Meteorological File
The model meteorological input file contains the meteorological data varying through time \((t)\) and requires the following variables:
- DATETIME – Datetime [yyyy-mm-dd hh:mm]
- T2 – Air temperature [°C]
- U2 – Wind speed [m s\(^{-1}\)]
- RH2 – Relative humidity [%]
- PRES – Atmospheric pressure [hPa]
- RRR – Precipitation [mm]
- N – Fractional cloud cover [0-1]
Alternatively, instead of using fractional cloud cover ( N ), the user can specify directly measured radiative fluxes:
- SWin – Shortwave radiation [W m\(^{-2}\)]
- LWin – Longwave radiation [W m\(^{-2}\)]
Optional Variables:
The following optional variables can also be used:
- T2_LAPSE – Air temperature lapse rate [°C m\(^{-1}\)]
(enables a temporally-variable lapse rate and overrides the 'air_temperature_lapse_rate' parameter) - D – Precipitation downscaling coefficient [-]
(hourly precipitation downscaling coefficient for the three-phase anomaly model) - PRECIPITATION_ANOMALY – Precipitation anomaly [-]
(annual precipitation variance / anomaly for the three-phase anomaly model)
An exemplar meteo CSV would therefore have the following format:
| DATETIME | T2 | U2 | RH2 | PRES | RRR | N |
|---|---|---|---|---|---|---|
| 2024-01-00 13:00 | 263.15 | 6.22 | 60.54 | 652.42 | 1.00 | 0.87 |
| 2024-01-00 14:00 | 264.56 | 8.71 | 66.22 | 653.18 | 2.50 | 0.95 |
| ⋮ | ⋮ | ⋮ | ⋮ | ⋮ | ⋮ | ⋮ |
| 2024-12-31 22:00 | 268.34 | 1.42 | 82.45 | 643.61 | 0.00 | 0.00 |
| 2024-12-31 23:00 | 269.21 | 2.20 | 81.56 | 644.22 | 0.00 | 0.00 |
Place the input CSV file in the data/meteo/CSV/ directory and then the `create_meteo_netcdf.py utility program can then convert it into NetCDF format.
The program is launched, by navigating to the utilities/create_METEO/ directory on the command line and executing the program create_meteo_netcdf.py with the following arguments:
- -c <meteo_csv>.csv – input meteo CSV file name
- -m <meteo_netcdf>.nc – output meteo NetCDF file name
- -s <yyyy-mm-dd hh:mm:ss> – (optional) start datetime
- -e <yyyy-mm-dd hh:mm:ss> – (optional) end datetime
cd utilities/create_METEO/
python3 create_meteo_netcdf.py -c <meteo_csv>.csv -m <meteo_netcdf>.nc
In Switzerland, hourly resolution meteorological data is readily available from a variety of stations on the Open Data platform of the Federal Office of Meteorology & Climatology (Meteo Swiss).
\((3)\) Illumination File
The model illumination input file determines whether grid nodes across the spatial domain \((x,y)\) are illuminated by the sun for any given timestep in a standard calendar and leap year \((t)\) :
The create_illumination_netcdf.py utility program can create this file from an existing static file.
The program is launched, by navigating to the utilities/create_ILLUMINATION/ directory on the command line and executing the program create_illumination_netcdf.py with the following arguments:
- -s <static_netcdf>.csv – input static NetCDF file name
- -i <illumination_netcdf>.nc – output illumination NetCDF file name
cd utilities/create_ILLUMINATION/
python3 create_illumination_netcdf.py -s <static_netcdf>.nc -i <illumination_netcdf>.nc
Attention
In order for the illumination file to be accurate, the static file should include all surrounding high-altitude terrain so that it can be determined whether they may obscure direct insolation onto the glacier. Glacier nodes (mask = 1) should also not be placed on the periphery of the static file.
Note
The illumination file is currently limited to a minimum of an hourly temporal resolution.