Soil Parent Materials
- Source
- Open Canada
- Published
- 2026
- Format
- multiple formats
- Access
- Open access
- License
- Open Government Licence
- Themes
- Soil Health, Machine Learning & AI, Remote Sensing & Satellites
About this dataset
The Soils Parent Materials raster prediction surface was crated using a Random Forest classifier trained with single component polygons from soils, terrain and ecosystem mapping projects, and 18 topographic predictor layers derived from the provincial 25m DEM. It is a prediction of mode of deposition of the soils parent material. The model was run by Ecoprovince and also enforced topographic constraints on some materials. The results of the Ecoprovince models were compiled into a provincial raster layer. The methodology and results are described in the research paper, Improved Soil Mapping in British Columbia, Canada with Legacy Soil Data and Random Forest Digital Soil Mapping Across Paradigms, Scales and Boundaries, 2016, ISBN: 978-981-10-0414-8 C. Bulmer, M. G. Schmidt, B. Heung, C. Scarpone, J. Zhang, D. Filatow, M. Finvers, S. Berch, S. Smith, DOI 10.1007/978-981-10-0415-5_24....
How the data was collected
The Soils Parent Materials raster prediction surface was crated using a Random Forest classifier trained with single component polygons from soils, terrain and ecosystem mapping projects, and 18 topographic predictor layers derived from the provincial 25m DEM.