|Adamchuk, Viacheslav -|
|Mat Su, A -|
|Ferguson, Richard -|
Submitted to: Transactions of the ASABE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: February 15, 2011
Publication Date: May 15, 2011
Citation: Adamchuk, V.I., Mat Su, A.S., Eigenberg, R.A., Ferguson, R.B. 2011. Development of an angular scanning system for sensing vertical profiles of soil electrical conductivity. Transactions of the ASABE. 54(3):757-767. Interpretive Summary: Agriculture is transitioning into an age of high technology; that change offers opportunities for better utilization of resources and a safer more-efficient supply of food for consumers. One issue faced by twenty-first century farmers is determining the extent and depth of soil variability. This work describes a potential solution of monitoring soil variability, both on the surface and deeper in the soil, all without disturbing the soil. A system that uses low frequency electromagnetic (EM) waves to measure soil conductivity was used at the heart of the project. The EM device was supported on a plastic sled by a rotating arm that was operated pneumatically. As the EM device rotated, it received signals that penetrated to various depths; these signals represented soil conductivity at different depths. A mature version of this system will provide producers with a means of mapping fields for judicious use of tillage and application of fertilizer, seed, and herbicides.
Technical Abstract: Apparent soil electrical conductivity (EC**a**) is typically mapped to define soil spatial variability within an agricultural field. Knowledge of the vertical variability of EC**a** is desired to define site-specific behavior of the soil profile. A Pneumatic Angular Scanning System (PASS) was developed to sense horizontal and vertical changes of EC**a** on-the-go using an electromagnetic induction (EMI) instrument using an angular scanning method. This sensor system consists of a sled with a rotating mechanism, an EMI sensor, an inclinometer, and a pneumatic actuator. The system was evaluated at the University of Nebraska-Lincoln Agriculture Research and Development Center (ARDC) near Mead, Nebraska. PASS was towed by an all-terrain vehicle (ATV) and operated from a field computer with specially designed data acquisition software. Rotation of the instrument allowed continuous transition between horizontal and vertical modes of operation. Nine discrete field locations with different soil conditions were used to compare PASS predictions with measurements obtained using a manual EC**a** probe. With two fixed depth layer assumption, R**2** was 0.91 for the linear regression between corresponding measured and predicted EC**a** values; and R**2** was 0.54 for the difference between the ECa of deep and shallow soil. Unfortunately, solving the system of linear equations for a more complex model of soil profile required inversion of an ill-conditioned (close to singularity) matrix, which was not feasible without regularization and an inversion procedure with non-negative constraint to be pursued in the future.