Interpretation of salinity characteristics of normal profile in estuarine region by using electromagnetic induction

Aiming at the problem of soil salinization in Yangtze River estuary area, normal salinity profile type extensively distributed in the area were studied. Multiple linear regression model between apparent electrical conductivity and soil electrical conductivity was established by using EM38 measurement and field sampling. Based on the analyses of soil salinity profile characteristics, Logistic model was propounded and fitted for all calibration salinity profiles, and the obtained Logistic model was then used to test the validation salinity profiles. The results showed that the horizontal value of electromagnetic apparent electrical conductivity was linearly correlated with the vertical value significantly. Both linear regression model and Logistic model exhibited high precision in simulated and predicted soil salinity. Prediction accuracy between them was statistically insignificant by the model accuracy test, but the Logistic model needed fewer parameters. The study suggested that the Logistic model proposed in this paper could not only reduce the amount of parameters but also had highly predicted precision. The research results can serve as a theoretical reference to the rapid and accurate assessment of soil salinity and the prevention of soil secondary salinization using electromagnetic induction.