Soil-Landscape Modeling and Remote Sensing to Provide Spatial Representation of Soil Attributes for an Ethiopian Watershed

Table 4

Root mean square error between field observations and predicted soil attributes for multiple regression and IDW and using different densities of observations.

Predicted soil attributes	R_180 (A)	I_180 (B)	R_60 (C)	I_60 (D)	B-A	D-C
Soil depth (cm)	26.4	32.64	33.7	43.37	6.24	9.67
Clay (%)	12.6	16.64	12.70	20.51	4.04	7.81
Silt (%)	7.3	10.25	8.59	14.64	2.95	6.05
Sand (%)	9.4	11.47	10.24	15.44	2.07	5.2
Organic matter (%)	1.39	1.53	1.55	1.82	0.14	0.27
Bulk density (g cm^−3)	0.18	0.25	0.24	0.36	0.07	0.12
pH	0.38	0.54	0.46	0.83	0.16	0.37
Total N (%)	0.29	0.46	0.09	0.34	0.17	0.25
Available P (mg kg^−1)	19.41	20.35	17.12	21.22	0.94	4.1
Surface stone cover (%)	12.2	14.53	14.15	20.62	2.33	6.47
Stone in the soil (%)	12.4	15.71	17.13	26.38	3.31	9.25

R_180 (A): root mean square error calculated from those predicted using 180 observations by multiple regression models, I_180 (B): root mean square error calculated from spatial interpolation (IDW: inverse distance weighted) using 180 observations, R_60 (C): root mean square error calculated from those predicted using 60 observations by multiple regression models, I_60 (D): root mean square error calculated from spatial interpolation (IDW: inverse distance weighted) using 60 observations, B-A: the difference in root mean square errors between those interpolated from 180 observations and those predicted using 180 observations by multiple regression models, and D-C: the difference in root mean square errors between those interpolated from 180 observations and those predicted using 180 observations by multiple regression models.