Estimation of soil physical properties from sensor-based soil strength and apparent electrical conductivity

Quantification of soil physical properties has traditionally been through soil sampling and laboratory analyses, which is time-, cost-, and labor-consuming, making it difficult to obtain the spatially-dense data required for precision agriculture. Soil strength and apparent electrical conductivity (EC_a) sensor measurements are significantly affected by soil texture, bulk density (BD), and water content (WC). The objective of this study was to explore the potential to estimate soil texture, BD, and WC using sensor-based soil strength and EC_a data obtained from sites with varying soil physical properties. Data collected from three research sites in Missouri included on-the-go horizontal soil strength at five depths up to 50 cm on a 10-cm interval, cone index at the same depths, EC_a measured by EM38 and Veris devices, and depth-dependent soil properties including soil texture, BD, and WC. Accuracy of the regression models for WC estimation varied, with better results when adding soil strength data to EC_a data and when including depth as a variable. Regressions estimating BD generally included EC_a, and also included soil strength at some depths and sites. BD estimates were improved considerably by adding WC to the model, suggesting the need for a WC sensor. Soil clay content (as a component of texture) was well-estimated by EC_a variables and the accuracy was improved in some cases by adding soil strength and depth variables. This study showed the potential of fusing data from multiple mobile proximal sensors to estimate major soil physical properties.