Table 12.
Information about the regional dataset used to evaluate the soil P distribution simulated in our study (GPASOIL-v1.1).
| Name | RMQS | LUCAS | STS |
|---|---|---|---|
| Reference | Jolivet et al.63 | Jones et al.19 | TFI64 |
| Website | Data can be download from: 10.15454/QSXKGA | Data can be downloaded after a request at ESDA website80: https://esdac.jrc.ec.europa.eu/content/lucas2015-topsoil-data | http://soiltest.tfi.org. |
| Spatial extent | France | Europe | Canada and USA |
| Years of sampling | 1st compaign of measurement (2002–2009) | 2015 | Different years were available in STS but we used the last year in common to GPASOIL-v1.1 and STS (i.e. 2015). |
| Soil layer | Value of the 1st soil layer considered was used (this soil layer is not necessary exactly equal to 0–30 cm but very close to it) | 0–0.2 m | Value for the top 0–0.3 m was computed by using the P concentration available for each soil layer within this horizon in STS. |
| Chemical extraction for P | Olsen P | Olsen P (information given in Tóth et al.81) | All measures were done using different P extraction method and then all data were converted to Bray and Kurtz P1 Equivalent. |
| Land-cover used to approach cropland and grassland | The categories “cropland” and “fruits” for the land-cover “occupation1” were used for our land-cover cropland. The category “grassland” was used for our land-cover “grassland”. | Categories “cropland” and “grassland” for the “main group of Land Cover” (LC0) were considered | Values are provided at county/province without distinction between cropland and grassland |
| Unit | gP2O5 (kg of soil)−1 | mgP (kg of soil)−1 | mgP (kg of soil)−1 |
| Treatments performed for comparison to our dataset |
- change in projection - average of all sites within 0.5 grid-cell (with average done separately for cropland and grassland) - conversion in kgP ha−1 by using the 0.5 resolution bulk density and volumetric fraction of coarse fragments used in our approach |
- average of all sites within 0.5 grid-cell (with average done separately for cropland and grassland) - conversion in kgP ha−1 by using the 0.5 resolution bulk density and volumetric fraction of coarse fragments used in our approach |
- depth average to compute 0–0.3 m value - conversion in kgP ha−1 by using the country-scale value of bulk density that we computed from the 0.5 resolution bulk density and volumetric fraction of coarse fragments used in our approach |