Enhancing understanding of soil electrical behavior: multi-variable analysis and correlation modeling

Abstract

Electrical resistivity (ER) approach leverages soil’s electrical behavior to detect its anomaly in a non-destructive (NDT) manner. This behavior is influenced by varying parameters which include soil’s resistivity, particle size gradation, physical parameters (water content, porosity and density), salinity and temperature which are amongst most crucial ones. Considering all above parameters, field ER tends to provide a continuous subsurface profile in a quick, effective and reliable manner resulting in increase of usage in the past decades. ER measurements are not only confined to in situ tests, but laboratory setup is also developed to estimate ER values. In this context, researchers in the past have adopted both field- and laboratory-based resistivity values to establish correlations with varying soil properties showcasing significant limitations in the form of site-specific nature with single input variable. In this context, the present study aims to evaluate effect of multiple parameters in the form of temperature, clay content, salinity, density, air content and water content on resistivity of soil. Results derived from the current analysis clearly showcases that effect of multiple parameters has a profound impact on soil’s resistivity compared to a single parameter. Further, correlationships were also developed involving simple and multiple regression analysis which resulted in formation of multi-variable model having coefficient of correlation (R²) value of 0.97. Furthermore, sensitivity of variables was also analyzed to ascertain effect of individual parameter on resistivity. Based on derived data, it could be inferred that density and temperature were found to be the most least sensitive input variables. Thus, results derived from the present study would play a pivotal role in accurate assessment of ER values.