Determining Soil Salinity from Soil Electrical Conductivity using Different Models and Estimates

Publisher Website
Google Scholar
Abstract
The appropriateness of two versions of a model describing electrical current flow in undisturbed soil was evaluated for purposes of diagnosing and mapping soil salinity (ECe). Different methods of measuring bulk soil electrical conductivity (ECa) and different ways of obtaining the values of other parameters required by the model were also evaluated using three different sets of data. The reliability of the variously predicted salinities were evaluated by comparing them against conventionally measured salinities using linear regression analyses, ranking tests, and a procedure based on the weighted sums of squared differences. It was found that salinity‐prediction accuracy, though relatively good, was underestimated for the instrumental/model techniques because of sample‐size differences and spatial‐variability effects. Conclusions arrived at from earlier sensitivity analyses about the suitability of the parameter‐estimation procedures were borne out by these results, i.e., that soil salinity appraisal and mapping can be quite adequately made using field measurements of ECa (by any one of three methods) and estimates of soil water content, bulk density, and surface conductance determined by “feel” methods.