8315df49-bde3-4138-8f71-9b722f3afd06 SoilGrids ISRIC – World Soil Information Laura Poggio Research Associate laura.poggio@wur.nl https://orcid.org/0000-0003-1892-0764 ISRIC – World Soil Information Luis M. de Sousa Geoinformatics Scientist luis.moreira.de.sousa@tecnico.ulisboa.pt https://orcid.org/0000-0002-5851-2071 ISRIC – World Soil Information Niels H. Batjes Senior soil science expert and coordinator WDC-Soils niels.batjes@isric.org https://orcid.org/0000-0003-2367-3067 ISRIC – World Soil Information Gerard B. M. Heuvelink Professor gerard.heuvelink@isric.org https://orcid.org/0000-0003-0959-9358 ISRIC – World Soil Information Bas Kempen Operations Manger bas.kempen@isric.org ISRIC – World Soil Information Eloi Ribeiro Geoinformatics eloi.ribeiro@isric.org https://orcid.org/0000-0002-2324-3000 ISRIC – World Soil Information David Rossiter Guest Researcher david.rossiter@isric.org https://orcid.org/0000-0003-4558-1286 SIB Swiss Institute of Bioinformatics Chiara Bortoluzzi Data Manager chiara.bortoluzzi@sib.swiss https://orcid.org/0000-0001-6589-6635 2021-06-14 English SoilGrids is a system for global digital soil mapping that uses state-of-the-art machine learning methods to map the spatial distribution of soil properties across the globe. SoilGrids prediction models are fitted using over 230,000 soil profile observations from the WoSIS database and a series of environmental covariates. Covariates were selected from a pool of over 400 environmental layers from Earth observation derived products and other environmental information including climate, land cover, and terrain morphology. The outputs of SoilGrids are global soil property maps at six standard depth intervals at a spatial resolution of 250 meters. Soil None Digital soil mapping The SoilGrids maps are publicly available under the CC-BY-4.0 Licence. Creative Commons https://creativecommons.org/licenses/by/4.0/ Mapserver entry page SoilGrids250m - WMS https://maps.isric.org/ Mapserver entry page SoilGrids250m - WCS https://maps.isric.org/ SoilGrids250m WebDAV https://files.isric.org/soilgrids/latest/data/ SoilGrids250m Google Earth Engine https://git.wur.nl/isric/soilgrids/soilgrids.notebooks/-/blob/master/markdown/access_on_gee.md SoilGrids250m Google Earth Engine https://git.wur.nl/isric/soilgrids/soilgrids.notebooks/-/blob/master/markdown/access_on_gee.md Global -180 180 84 -56 ISRIC - World Soil Information SoilGrids Team soilgrids@isric.org SoilGrids uses state-of-the-art statistical methods for digital soil mapping, relying exclusively on open source tools. The models are tailored per soil property and fitted using documented models. For each property, a global model is calibrated using a spatially stratified 10-fold cross-validation procedure. The model produces values at each map location (cell) and standard depth. The prediction distribution is captured in four different maps reporting its 5%, 50%, and 95% quantiles, and the mean. The mean represents the 'expected value' and provides an unbiased prediction of the soil property, whereas the median yields that value for which there is a 50% probability that the true soil property value is greater and a 50% probability that the true value is smaller. The 0.05 and 0.95 quantiles present the lower and upper boundaries of a 90% prediction interval and may be used as a measure of prediction uncertainty. This interval presents a value range that contains the true soil property value for each cell (which one would measure from a soil sample taken at the centre of the cell) with 90% probability. Quantiles of the distribution were computed with Quantile Regression Forests as implemented in the ranger package in R. The mean was computed using the default random forests algorithm. All elements have the same name, a triplet separated by underscores: property_depthInterval_quantile. For instance, to download the 5%-quantile prediction of coarse fragments in the 5 cm depth interval, the user must fetch the files 'cfvo_5-15cm_Q05.vrt' and 'cfvo_5-15cm_Q05.ovr' plus the 'cfvo_5-15cm_Q05' folder. Poggio, L., de Sousa, L. M., Batjes, N. H., Heuvelink, G. B. M., Kempen, B., Ribeiro, E., and Rossiter, D.: SoilGrids 2.0: producing soil information for the globe with quantified spatial uncertainty, SOIL, 7, 217–240, 2021. https://doi.org/10.5194/soil-7-217-2021 SoilGrids is a system for global digital soil mapping that uses state-of-the-art machine learning methods to map the spatial distribution of soil properties across the globe. SoilGrids prediction models are fitted using over 230,000 soil profile observations from the WoSIS database and a series of environmental covariates. Covariates were selected from a pool of over 400 environmental layers from Earth observation derived products and other environmental information including climate, land cover, and terrain morphology. The outputs of SoilGrids are global soil property maps at six standard depth intervals at a spatial resolution of 250 meters. SoilGrids is developed and maintained by a team of scientists from the International Soil Reference and Information Centre in Wageningen, the Netherlands SoilGrids250m VRT, OVR, GeoTIFF bdod Bulk density Bulk density of the fine earth fraction kilogramPerCubicDecimeter cec Cation exchange capacity Cation exchange capacity of the soil cmol(c)/kg cfvo Coarse fragments Volumetric fraction of coarse fragments (>2 mm) cm3/100cm3 (vol%) clay Clay Proportion of clay particles (more than 0.002 mm) in the fine earth fraction g/100g (%) nitrogen Total nitrogen Total nitrogen (N) gramPerKilo ocd Organic carbon density kilogramPerMeterCubed ocs Organic carbon stock kilogramPerMeterCubed phh2o Soil pH pH sand Sand Proportion of sand particles (>0.05/0.063 mm) in the fine earth fraction g/100g (%) silt Silt Proportion of silt particles (higher or equal to 0.002 mm and lower or equal to 0.05/0.063 mm) in the fine earth fraction g/100g (%) soc Soil organic carbon Soil organic carbon content in the fine earth fraction gramPerKilo