Accurate characterization of soil properties such as soil water content (SWC) and bulk density (BD) is vital for hydrologic processes and thus, it is importance to estimate θ (water content) and ρ (soil bulk density) among other soil surface parameters involved in water retention and infiltration, runoff generation and water erosion, etc. The spatial estimation of these soil properties are important in guiding agricultural management decisions. These soil properties vary both in space and time and are correlated. Therefore, it is important to find an efficient and robust technique to simulate spatially correlated variables. Methods such as principal component analysis (PCA) and independent component analysis (ICA) can be used for the joint simulations of spatially correlated variables, but they are not without their flaws. This study applied a variant of PCA called independent principal component analysis (IPCA) that combines the strengths of both PCA and ICA for spatial simulation of SWC and BD using the soil data set from an 11 km2 Castor watershed in southern Quebec, Canada. Diagnostic checks using the histograms and cumulative distribution function (cdf) both raw and back transformed simulations show good agreement. Therefore, the results from this study has potential in characterization of water content variability and bulk density variation for precision agriculture.
