yquilcaille / fwi_cmip6 Goto Github PK

The Canadian Fire Weather Index (FWI) is a commonly used index to calculate and assess the fire weather. Several packages have proposed different adjustments to the initial algorithm of the FWI. We have gathered these improvements to provide an algorithm that can be used globally. This code has been used to calculate the FWI using all available simulations of the Climate Model Intercomparison Project Phase 6 (CMIP6). The daily FWI has then been summarized into four annual indicators: (i) maximum value of the FWI (fwixx), (ii) number of days with extreme fire weather (fwixd), (iii) length of the fire season (fwils), and (iv) seasonal average of the FWI (fwisa). The corresponding publication is available at https://doi.org/10.5194/essd-2022-413 while the database is available at https://doi.org/10.3929/ethz-b-000583391.

When using this code, please acknowledge the publication (https://doi.org/10.5194/essd-2022-413).

These scripts are developed in Python 3.7. Instructions on how to use the code and required packages are provided below. Any feedback and contributions are welcome. This code comes into several components:

• Computation of the FWI:
  • "_calc_FWI-CMIP6.py": calculates the FWI on CMIP6 runs.
  • "functions_calc_FWI.py": functions used for calculation of the FWI. The original algorithm for the FWI (Simard, 1970: https://cfs.nrcan.gc.ca/publications?id=36178) and has been updated in several codes. The basis of this code is pyfwi (https://github.com/buckinha/pyfwi). It has been adapted to handle arrays, not only scalars. Besides, some exceptions are now handled. The code has been extended by integrating improvements from NCAR (https://github.com/NCAR/fire-indices) and from cffdrs (https://rdrr.io/rforge/cffdrs).
  • "functions_load_CMIP6.py": functions used to load CMIP6 data.
  • "functions_support.py": functions used throughout components.
  • "available_files": repository that gather information on the runs. In particular, CMIP6 runs may have to undergo corrections on their reported grids. Worse cases, some CMIP6 runs have to be excluded.
  • "_run_calc_FWI-CMIP6.sh": simply a script to run an ensemble of jobs "_calc_FWI-CMIP6.py" on a server.
• Regrid the FWI: the FWI from CMIP6 were regridded to a single grid using the CMIP6-ng processor (https://git.iac.ethz.ch/cmip6-ng/cmip6-ng), although adapted to integrate the variable FWI, which was not in CMIP6. Nevertheless, the following step can be performed on the native grid of the outputs without regridding. Otherwise, the user can use its own favorite tool for regridding (e.g. CDO).
• Annual indicators of the FWI:
  • "_extract_annual_indicators.py": calculates annual indicators of the FWI from daily FWI.
  • "functions_calc_indicators.py": functions used for the calculation of annual indicators.
• Mask for infrequent burning:
  • "_mask_infreq_burning.py": calculates the mask for chosen grids.
• Plots of FWI on CMIP6 data:
  • "_plots_FWI.py": plots used for representation of the FWI in CMIP6 runs (DOI coming soon).
  • "functions_support_plot.py": functions used for the plots.

Requirements on packages:

• cartopy (0.20.2)
• cftime (1.6.1)
• matplotlib (3.5.2)
• numpy (1.23.1)
• pandas (1.4.3)
• regionmask (0.9.0)
• seaborn (0.11.2)
• xarray (2022.6.0)