A model for soil moisture content prediction based on the change in ultrasonic velocity and bulk density of tillage soil under alternating drying and wetting conditions

Abstract The feasibility of introducing soil bulk density as a factor to distinguish different levels of soil compactness for the predictive modeling of soil moisture content was verified. Outdoor rainfall was simulated to set different initial soil moisture contents, after which the soil was allowed to air dry in a natural environment. Four drying and wetting (D-W) cycles were simulated in the experiment. An ultrasonic detection device was employed to measure the soil ultrasonic velocity and calculate the corresponding soil bulk density. An adaptive weighted data fusion algorithm was adopted to integrate the soil ultrasonic velocity and bulk density results, and a bivariate linear regression model was established to describe the relationships of soil ultrasonic velocity and bulk density with soil moisture content over D-W cycles. The model suitably described the soil moisture content in a given cultivated layer as a function of soil ultrasonic velocity and bulk density.