Skip to content

Latest commit

 

History

History
70 lines (51 loc) · 2.94 KB

example-workflows.md

File metadata and controls

70 lines (51 loc) · 2.94 KB
Raw

Example workflows

This document gives some examples of workflows for compiling input data, running the model and working with output data. Example shell scripts are provided to give an idea of how to implement these workflows, but these will need to be adjusted to work on your system.

First step: Compile the model

If you haven't already, you need to compile the model to an executable file. This only needs to be done once (unless the model code changes), even if using different data. Instructions are given in the README.md.

Compiling/editing input data > running the model

  • Use the NanoFASE data module to compile or edit input data. Refer to this README.md for documentation on how to do this.
  • The data compilation will produce two files: A NetCDF file and a Fortran Namelist (.nml) file.
  • Create a model config file (example given here: config.example/config.example.nml) or update an existing one, ensuring that &data > input_file points to the NetCDF file, and &data > constants_file points to the Fortran Namelist file.
  • Run the model, as detailed in the README.md.

Example: updating emissions scenarios

Say you already have input data files in the data directory (all paths relative to the repo root):

data/
|-- data.nc
|-- constants.nml
|-- new_emissions_areal_pristine_soil.tif

You wish to edit the NetCDF file (data.nc) to include a new raster for areal emissions of pristine nanomaterials to soil (new_emissions_areal_pristine_soil.tif). The relevant variable in the NetCDF file is emissions_areal_pristine_soil and we can write an edit config file for the NanoFASE data module to do this:

input_nc_file: ./data/data.nc
output_nc_file: ./data/data_edited.nc
emissions_areal_soil_pristine:
  type: raster
  path: ./data/new_emissions_areal_pristine_soil.tif
  units: kg/m**2/year

Save this to vendor/nanofase-data/config.edit.yaml. Now we can run the data compiler, which is provided as a submodule to this repo:

$ vendor/nanofase-data/nanofase_data.py edit vendor/nanofase-data/config.edit.yaml

After this has finished, our data folder should have the new NetCDF file:

data/
|-- data.nc
|-- data_edited.nc
|-- constants.nml
|-- new_emissions_areal_pristine_soil.tif

To run the NanoFASE model, a config file is need (see README.md). Assuming you already have one located at config/config.nml, edit the &data group of this file to point to the new data:

&data
input_file = "data/data_edited.nc"
constants_file = "data/constants.nml"
output_path = "data/output/"
...

Now you're ready to run the model with this updated config file:

$ make run CONFIG_PATH=config/ CONFIG_FILE=config.nml

The output data will be saved at data/output (specified in the config file, as per above).