How to determine a representative and economical soil sampling method that combines soil properties with advanced technologies has been an unsolved issue in soil related studies. This paper proposes a new method for soil sampling based on some measured salinity data and remote sensing images, as well as analysis of spatial distributions of soil properties in Lubotan land reclamation area in Shaanxi, China. The results showed that with the available data sets of 33 points, up to 101 unknown points can be estimated, and further interpolation of 343 points displayed spatial distribution of soil salinity in the study area, the coefficients of determination (R2) for the predicted soil salinity were 0.817, respectively. A further Kriging analysis for top soil salinity distribution in the study area showed that soil salinity has a medium degree of autocorrelation and low variability. The study in this paper may help understand the effect of soil reclamation efforts and local water management practice.