Maybe, this could be added to the SoilDataPrep documentation - some meta data on the TAXNWRB soil classes:

SoilGrids IDs and classes:

• Codes and names:

ftp://ftp.soilgrids.org/legends/TAXNWRB.qml
(explanation: SoilDataPrep creates a tif-map called "soils_catchment", with the most probable soil class; only values are assigned to each pixel. Here you can find, which value is which FAO soil class.)

• Legend & short characteristic of soil types:

ftp://ftp.soilgrids.org/data/recent/SoilGrids250m_TAXNWRB_legend_June_2016.pdf

• Index of all SoilGrids legends:

ftp://ftp.soilgrids.org/legends/

Explanation & paper on SoilGrids:

• General explanation SoilGrids:
  https://www.isric.org/explore/soilgrids/
• SoilGrids250m:
  Global gridded soil information based on machine learning:
  https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0169748#pone.0169748.ref015

FAO legend & explanation of soil types:

• World reference base for soil resources 2014:
  http://www.fao.org/3/i3794en/I3794en.pdf

The new SoilGrids release is not accessible anymore with SoilDataPrep.
(https://www.isric.org/explore/soilgrids)

Data access is now possible via:

• WMS
  (https://maps.isric.org
  GIS-example: https://www.isric.org/instruction-wms)
• WCS
  (https://www.isric.org/web-coverage-services-wcs
  R-example: https://git.wur.nl/isric/soilgrids/soilgrids.notebooks/-/blob/master/markdown/wcs_from_R.md)
• WebDAV
  (https://files.isric.org/soilgrids/latest/data/
  R-example: https://www.isric.org/explore/soilgrids/soilgrids-access)

Map projection seems to have changed to "Homolosine" which might require a reprojection for users' input maps DEM and subbasin-shapefile (previously given in WGS84 LatLon).

I did not yet check the layer names for changes. The new names are listed here:
https://files.isric.org/soilgrids/latest/data/

Also, Soil Organic Carbon has been calculated differently in the latest release.
(See: https://www.isric.org/explore/soilgrids/faq-soilgrids#What_happened_to_the_maps_of_soil_types
-> Are there differences between the soil organic carbon stocks in the latest and former versions of SoilGrids?)

Check if the correct soilgrids layer depts ( #Thickness [mm] - set thickness of soil layers from those used in soilgrids
soil_means2$sl1thickness = 25
soil_means2$sl2thickness = 75
soil_means2$sl3thickness = 120
soil_means2$sl4thickness = 230
soil_means2$sl5thickness = 350
soil_means2$sl6thickness = 700
soil_means2$sl7thickness = 500
) are correctly exported. In one of Anne's parameterizations this was not the case.