S-Function / soil erodibility / dust source function for mineral dust aerosol emissions, including high latitude sources

Datasets and scripts for the Ginoux S-Function (Ginoux et al., 2001, https://doi.org/10.1029/2000JD000053) used to calculate mineral dust aerosol emissions. The S-function (also sometimes called the soil erodibility or mineral dust aerosol source function) represents the potential emission strength of mineral dust aerosol from bare soils, scaled by a factor that represents enhanced emissions from topographical depressions. Here, it is calculated at 60 arc-seconds from the ETOPO2022 and the GLCNMOv3 datasets, using a spatial scale of 10.0°x10.0° to define hydrological basins, following the methodology of Groot Zwaaftink et al. (2016, https://doi.org/10.1002/2016JD025482) Data in Iceland has been replaced by soil erodibility data from https://doi.org/10.3390/atmos13091345 (digitized from Figure 2) References For detailed presentation and discussion of this dataset, see - Marelle et al.: Modeling mineral dust aerosols in the Arctic with the regional model WRF-Chem 4.6.1, 2026, in preparation for GMDFor the input datasets: - ETOPO2022: http://doi.org/10.25921/fd45-gt74 - GLCNMOv3: https://doi.org/10.5539/jgg.v9n3p1 and https://github.com/globalmaps/gm_lc_v3 - Geospatial Information Authority of Japan, Chiba University and collaborating organizations - Icelandic erodibility data: https://doi.org/10.3390/atmos13091345 Contents In data/- sfunc_60s.nc: The S-function at 60 arc-seconds, replaced by data from erodibility_iceland_latlon.nc in Iceland. Best for regional applications, dx > 10 km- sfunc_0_1deg.nc: sfunc_60s.nc regridded to 0.1°x0.1°. Best for global applications, dx > 100 km- sfunc_0_5deg.nc: sfunc_60s.nc regridded to 0.5°x0.5°. OK for global visualization- erodibility_iceland_latlon.nc: Soil erodibility in Iceland, from https://doi.org/10.3390/atmos13091345 Figure 2 These datasets can be re-created using the provided python3 scriptsIn src/- create_glcnmo_landmask.py: prepares the GLCNMOv3 landcover files and calculates a fraction of bare soil and a fraction of water in each grid cell- calculate_sfunc.py: calculates the S-function and creates sfunc_60s.nc- regrid_sfunc.sh: regrids the S-function from 60s to 0.1 and 0.5 degree resolutions, using CDO (Climate Data Operators), creates sfunc_0_1deg.nc and sfunc_0_5deg.nc In wrfchem_dust_preprocessor/- regrid_wrfinput_erod.py : regridding script to interpolate the S-function to the WRF-Chem model grid. It uses conservative remapping in CDO to regrid the S-function to variable EROD_HL in the WRF-Chem wrfinput file. To use: python regrid_wrfinput_erod.py ./wrfinput_d01 'COMMENT' Compatible with WRF-Chem-Polar https://github.com/WRF-Chem-Polar/WRF-Chem-Polar- create_cdo_wrfgridfile.py: used by regrid_wrfinput_erod.py to create the CDO WRF grid description file for remapping- wrf_utilities.py: used by create_cdo_wrfgridfile.py to do map projections to WRF grid