Table 3.

SLUD driver and its representation in (data = GPASOIL-v0) and (data = GPASOIL-v1).

Driver name	SLUD (P in sludges)
Description of its effect on the soil P	P in sludges from sewage treatment that are spread on cropland soils
Corresponding variables in our approach	$f{P}_{slud}^{out\to m}$ with m in {i-lab, o-lab, o-sta}),
Dataset ref used in (data = GPASOIL-v0)	Not considered in Ringeval et al.11
Issues related to the use of this dataset in GPASOIL-v0
Representation in (data = GPASOIL-v1)	$f{P}_{slud}^{out\to x-tot}$ was estimated by combining human P excretions with the fraction of sewage sludge that is treated and the P removal efficiency of treatments plants following van Puijenbroek et al.52 and Demay et al.9. Human excretions were computed by using excretion rates and human population. Computations were performed at country-scale then all crop within the same country receive the same P (in kgP ha−1). For the composition of P in sludges, we used the same labile vs stable contribution as for manure. See Eq. 56 and Text S5.
Characteristics of the spatially explicit dataset used in (data = GPASOIL-v1)	Human P excretion rates for big World regions, as well as fraction of sewage sludge that is treated, both at country scale, were provided by van Puijenbroek et al.52. Human P excretion rates were available for the years 1970 and 2010 while the fraction of sewage sludge that is treated was available for years 1990, 2000, 2010. Human population was provided by HYDE 3.223 and Area at country scale was computed from the dataset used in LUCC (Table 5). The corresponding between big World regions and iso country code was provided by the IMAGE framework region classification (https://models.pbl.nl/image/index.php/Region_classification_map).