Abstract: Taking into account the problem of salinization existing in Lower Yellow River Delta, with the application of electromagnetic induction instruments (EM38 and EM31) and their mobile sensing system, response relation between apparent soil electrical conductivity and soil salinity in typical field of the Yellow River Delta was analyzed by using GIS and geostatistics, then the spatial variability of soil salinity characterized by apparent soil electrical conductivity was discussed and spatial distribution of soil salinity was further evaluated quantitatively. Results indicated that apparent soil electrical conductivity was well correlated with salinity, and interpretation accuracy of soil salinity at different layers could be improved when adopting the optimal combination of EM38h and EM31h. Apparent soil electrical conductivity EM38h and EM31h both exhibited moderate spatial variation, and nested spherical models were fitted for semivariance of EM38h and EM31h because of the compound scale effect of their spatial structures. Spatial autocorrelation of EM31h was better than that of EM38h, and short-range variation was the chief constitute of spatial heterogeneity for both EM38h and EM31h. Spatial distribution pattern of soil salinity was directly controlled by structural factors, while stochastic factors were of importance for the formation of spatial distribution patterns. Salinization classification showed that heavy salinized soil was the predominant soil type as a whole and soil salinity exhibited the trend of top enrichment in the study area. The predominant soil types were saline soil and heavy salinized soil while the distribution area of desalinized soil was negligible at all layers. The research results provided a reference for the survey and quality evaluation of soil properties at different scales ranging from farms to regional levels, and it could be used as the guide for the efficient utilization of agricultural water-soil resources.

Key words: electromagnetic induction instrument, soil salinization, spatial distribution, quantitative evaluation, Yellow River Delta