South Africa

Napo Monyane 1826690@students.wits.ac.za I used the SOTER_ZA soil database and can share it if requested.

Best option: iSDAsoil (Hengl et al., 2021)

This is the most suitable dataset for your use case. The iSDAsoil dataset provides soil texture classes derived from sand, silt, and clay fractions at 30 m resolution, for 0–20 cm and 20–50 cm depth intervals. It's projected in WGS84 and available as Cloud-Optimized GeoTIFFs (COG). Predictions were generated using multi-scale Ensemble Machine Learning with 250 m covariates (MODIS, PROBA-V, climate) and 30 m covariates (DTM derivatives, Landsat, Sentinel-2).

The classification assigns each pixel to one of twelve USDA texture classes, with continuous fraction predictions (sand, silt, clay) achieving concordance correlation coefficients of 0.78–0.85. arxiv

It covers all of Africa including South Africa. Download is available via Zenodo: https://zenodo.org/records/4094616

Alternative: Africa SoilGrids Texture (ISRIC, 2015)

This dataset provides USDA texture classes at 6 depth intervals, derived from sand, silt, and clay contents predicted using the Africa Soil Profiles Database (AfSP) v1.2, at 250 m resolution. It covers all 12 standard USDA classes: clay, silty clay, sandy clay, clay loam, silty clay loam, sandy clay loam, loam, silty loam, sandy loam, silt, loamy sand, and sand. Available from ISRIC at https://data.isric.org ISRIC

Also worth knowing: OpenLandMap via Google Earth Engine

OpenLandMap provides USDA soil texture classes at 250 m resolution for 6 depth intervals (0, 10, 30, 60, 100, and 200 cm), accessible directly through Google Earth Engine. This is very convenient for clipping to a South Africa extent without downloading the full Africa-wide rasters. Google

Practical recommendation for MIKE SHE:

For a MIKE SHE project in South Africa, I'd suggest iSDAsoil at 30 m as your primary source — the resolution is excellent for catchment-scale modeling, the USDA classification aligns directly with UNSODA lookup tables, and the continuous sand/silt/clay fractions (rather than just the categorical class) are available separately, which gives you more flexibility for parameterization. You'd then cross-reference with UNSODA to get the hydraulic parameters (Ksat, θs, θr, etc.) per texture class.

Would you like help setting up the texture-to-UNSODA mapping table, or with extracting/processing the iSDAsoil data for a specific South African catchment?