Spatial prediction of Coarse fragments in Switzerland at 30 m resolution

The Coarse Fragment (CF) maps were developed using a quantile random forest machine learning approach. The model integrates remotely sensed and derived environmental covariates including terrain, climate, vegetation, and soil properties. The maps are provided in GeoTIFF format at a 30 m spatial resolution for four soil depths: 0, 30, 60, and 100 cm. The unit is percent by volume. The methodology follows Gupta et al. (2024). Based on spatial five-fold cross-validation, the model performance was: MEC: 0.18 CCC: 0.32 RMSE: 8.3 Bias: 0.56 Among all predictors, soil depth was the most important covariate, followed by topography, climate, and lithology. Uncertainty was assessed using quantile regression. However, because coarse fragment values are often close to zero, the uncertainty estimates may be less reliable (Poggio et al., 2021). Therefore, these maps should be used with caution. Other soil property maps for Switzerland including soil organic carbon (SOC), clay, sand, nitrogen, and phosphorus can be downloaded here: https://doi.org/10.5281/zenodo.7821649 The bulk density map for Switzerland can be downloaded from: https://doi.org/10.5281/zenodo.15274427 References Gupta, S., Hasler, J. K., & Alewell, C. (2024). Mining soil data of Switzerland: New maps for soil texture, soil organic carbon, nitrogen, and phosphorus. Geoderma Regional, 36, e00747. Poggio, L., De Sousa, L. M., Batjes, N. H., Heuvelink, G. B., Kempen, B., Ribeiro, E., & Rossiter, D. (2021). SoilGrids 2.0: producing soil information for the globe with quantified spatial uncertainty. Soil, 7(1), 217–240.