Spatial prediction of some physicochemical soil properties and magnetic susceptibility using digital topographic model

This research was performed to study the spatial prediction of physicochemical soil properties and magnetic susceptibility (MS) using digital topographic model in Ardal district, located in Chaharmahal and Bakhtiari province. The selected soil variables for this research included organic matter content, nitrogen content, electrical conductibility, particle size distribution and magnetic susceptibility measured at 90 points on a regular 20m grid. The digital elevation model (DEM) of the study area was prepared in 3×3m pixels and the primary and secondary topographic attributes were calculated using DEM. The results showed that soil properties were normally distributed as confirmed by the Kolmogorov-Smirnov test and the values on skewness. The CVs were high for sand (0.52) and for Bd (0.1). Finally, multiple regression analyses relating physicochemical soil properties and magnetic susceptibility and topographic attributes were performed and then model were validated by additional sample points (18 of 90). The developed regression models showed significant relationships between soil properties and topographic attributes such as wetness index, slope, elevation and soil erodibility factor. The mean errors and root mean square errors in the validation of the models were low and somewhat acceptable. The regression equation could explain 20 to 67% of the total variability measured in the selected soil properties in the study area. This research indicated that using of topographic attributes is a powerful tool in soil properties prediction.