REPEATABILITY OF SOIL APPARENT ELECTRICAL CONDUCTIVITY MEASURED BY A COULTER SENSOR

Abstract

Apparent electrical conductivity (ECa) measured using an on-the-go coulter sensor offers advantages for mapping soil variability because detailed data can be collected easily and inexpensively using on-the-go ECa sensors. However, there has been little research investigating the repeatability of these sensors, which may be defined as their ability to reproduce the same ECa measurement when operated in the same location under the same operating and field conditions. If the output of the coulter ECa sensor is not repeatable, the accuracy and reliability of the resulting maps and management decisions would be compromised. Therefore, the objective of this study was to evaluate the repeatability of the coulter sensor by comparing ECa data from two passes in barley stubble at two 1.6-ha sites, one with a sandy loam soil texture (Nesson site) and the other, a clay loam soil texture (Montana State University Eastern Agricultural Research Center site). Sampling points were approximately 1.45 m apart in the direction of travel for both passes. The ECa measurements from both passes were compared at shallow (0-30 cm) and deep (0-90 cm) soil depths. The coefficients of variation of ECa measurements for shallow and deep depths from pass 1 were higher than those from pass 2 at both sites. The root mean square error values of ECa measurements between pass 1 and pass 2 at shallow and deep depths for the Nesson site were 0.76 and 0.51 mS m -1 , respectively, whereas the root mean square errors for the Montana State University Eastern Agricultural Research Center site were 4.06 and 2.93 mS m -1 at shallow and deep depths, respectively. The repeatability was evaluated using a 95% confidence interval for the differences between ECa measurements of the two passes. Results demonstrate marginally acceptable repeatability between the two passes at shallow depths and acceptable repeatability at deep depths. The reasons for lack of agreement between pass 1 and pass 2 in ECa measurements at shallow depths could have resulted from soil disturbance and compaction caused by the coulter sensor during the pass 1 process. Regardless of discrepancies for shallow depths, the results indicate that the on-the-go ECa sensors can be useful and provide reliable data for describing field spatial variability in precision farming. This study was conducted to represent field conditions under which this equipment will likely be used, and further work is needed to confirm the repeatability of the coulter at shallow depths.