Table 13 Spearman correlations computed between the different dataset (RMQS, LUCAS, STS) and Pi-lab simulated by McDowell et al.58, Zhang et al.12, GPASOIL-v0 or GPASOIL-v1.1.

From: A global dataset on phosphorus in agricultural soils

Dataset used for the comparison

Spatial extent and years of sampling

Modeling approach

Years used in the modeling approach to compare with the dataset

Spearman correlation over cropland

Spearman correlation over grassland

Spearman correlation over cropland + grassland

RMQS

France 2002–2009

McDowell et al.58

2009

+0.33 (p-value = 0.00)

+0.36 (p-value = 0.00)

 

Zhang et al.12

2010

−0.09 (p-value = 0.27)

  

GPASOIL-v0

2005

+0.03 (p-value = 0.77)

−0.11 (p-value = 0.27)

 

GPASOIL-v1.1

2002–2009

+0.01 (p-value = 0.94)

+0.32 (p-value = 0.00)

 

LUCAS

France 2015

McDowell et al.58

2009

+0.39 (p-value = 0.00)

+0.56 (p-value = 0.00)

 

Zhang et al.12

2010

+0.01 (p-value = 0.92)

  

GPASOIL-v0

2005

+0.04 (p-value = 0.68)

−0.22 (p-value = 0.02)

 

GPASOIL-v1.1

2015

−0.01 (p-value = 0.89)

+0.07 (p-value = 0.34)

 

LUCAS

Europe 2015

McDowell et al.58

2009

+0.56 (p-value = 0.00)

+0.58 (p-value = 0.00)

 

Zhang et al.12

2010

−0.07 (p-value = 0.03)

  

GPASOIL-v0

2005

+0.05 (p-value = 0.11)

−0.10 (p-value = 0.01)

 

GPASOIL-v1.1

2015

+0.18 (p-value = 0.00)

−0.13 (p-value = 0.00)

 

STS

Canada and USA 2015

McDowell et al.58

2009

  

−0.20 (p-value = 0.35)

GPASOIL-v0

2005

  

−0.23 (p-value = 0.35)

GPASOIL-v1.1

2015

  

+0.32 (p-value = 0.03)

  1. Bold front is used for p-value ≤ 0.05.