Skip to main content
ARS Home » Midwest Area » Columbia, Missouri » Cropping Systems and Water Quality Research » Research » Publications at this Location » Publication #273552

Title: Modeling soil electrical conductivity-depth relationships with data from proximal and penetrating ECa sensors

item Sudduth, Kenneth - Ken
item Myers, David
item Kitchen, Newell
item Drummond, Scott

Submitted to: Geoderma
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/8/2012
Publication Date: 3/26/2013
Citation: Sudduth, K.A., Myers, D.B., Kitchen, N.R., Drummond, S.T. 2013. Modeling soil electrical conductivity-depth relationships with data from proximal and penetrating ECa sensors. Geoderma. 199:12-21.

Interpretive Summary: Apparent soil electrical conductivity (ECa) has become widely used to map within-field soil variability. The main soil parameters affecting ECa vary from place to place, and maps made with commercial mobile ECa sensors have been related to variations in soil parameters such as salinity, texture or moisture. The resulting maps have then been used to guide site-specific management such as fertilizer application or seeding rate. Additionally, by combining data from multiple mobile ECa sensors and applying various mathematical and statistical techniques, it is possible to infer by-depth variations in soil properties. This information is particularly useful for understanding the dynamics of subsurface water movement across landscapes. However, existing analysis approaches are not always reliable due to the mathematical structure of the ECa datasets. Our goal in this research was to overcome this limitation by combining mobile ECa sensor data with point measurements of soil layer conductivity obtained using an ECa-equipped penetrometer. We found that the penetrometer ECa data improved our ability to create a good mathematical model of the soil profile and also provided an efficient way to obtain calibration data for the model. On a test dataset our procedure was better able to represent variations of ECa in the soil profile than previous methods. After further validation this new procedure could allow users of ECa sensors to obtain more accurate estimates of how soil properties vary and provide better information on which to base management decisions.

Technical Abstract: Apparent soil electrical conductivity (ECa), a widely used proximal soil sensing technology, is related to several important soil properties, including salinity, clay content, and bulk density. Particularly in layered soils, interpretation of ECa variations would be enhanced with better calibrations to depth-wise variations in these soil properties. Thus, the objective of this research was to combine point measurements of layer conductivity obtained using ECa -equipped penetrometers with mapped ECa data from proximal sensors for improved quantification of conductivity-depth relationships in layered soils. Data were collected from a set of large plots on a claypan-soil landscape central Missouri which had been managed in grain and perennial grass cropping systems for 20 years. Soil ECa variation with depth was represented by both two- and three-layer models that were visualized and parameterized with the use of penetrometer ECa data. The three layer model provided a more realistic representation of the soil profiles in the study area and provided similar accuracy as the commonly used two layer model. Penetometer ECa data also provided efficient and accurate (r2=0.92) estimation of calibration-point topsoil depth as an alternative to soil coring and manual determination. A key to accurate model calibration was selection of calibration points from areas of spatially-homogeneous proximal ECa. Model results more closely approximated measured penetrometer ECa when separate models were calibrated and applied to grain and perennial grass plots. Good model estimates were possible when using data from as few as six to eleven points for model calibration. Combining penetrometer ECa with proximal ECa data improved modeling of conductivity-depth relationships in terms of model selection, model parameterization, and model calibration.