Search

Comparison of IDW and Kriging interpolators in the spatial variation of soil, pH, Ca, CICE, and P. The objective of this study was to compare the Kriging and IDW (Inverse Distance Weighting), as the most frecuently-used interpolators in spatial analysis studies. From a 2 467 m2 area, 61 soil samples in a quadric design were taken at 0-15 cm depth. The average distance between sample points was 3.5 m. From the chemical analysis, only the variables pH, Ca, CICE, and P were used for the interpolations. Every single soil sample point was georeferenced. The GS+ for Windows program was used for variogram analysis, interpolation and cross validation. The spherical model was preferred on almost all the variograms. For comparison purposes, a cross validation was performed. As accuracy measurements, the Mean Absolute Error (MAE) and the Mean Square Error (MSE) were calculated. Also as a goodness-of-fit measurement, the prediction (E) was obtained, in order to compare both interpolators. Both procedures were satisfactory, however the Kriging method was better for pH, Ca, and CICE predictions, while IDW was better for P prediction. [ABSTRACT FROM AUTHOR]