Evaluation of Electromagnetic Induction as a Noninvasive Technique for Monitoring Water Movement into and Beneath Waste Disposal Facilities

The purpose of this study was to evaluate the use of electromagnetic induction to noninvasively monitor water content in waste disposal facility cover soils. We compared apparent electrical conductivity measurements monitored with the EM38 ground conductivity meter with water content monitored with a neutron probe at 20 locations over an 18-month period from August 1998 to January 2000. Two cover designs were monitored: a gel/asphalt barrier at 1.3 m depth and a capillary barrier at 2.0 m depth. The EM38 instrument was operated in both the vertical and horizontal dipole modes with the instrument resting on the ground surface and all data were normalized to 25°C. Linear regression techniques were applied to analyze the survey data. Water content to a depth of 0.75 m was correlated with horizontal dipole mode data and water content to depths of 1.1 m and 1.5 m was correlated with vertical dipole mode data. Initially higher water content values decreased by an average of 0.10 m^3/m^3 in the top 0.75 m and an average of 0.07 m^3/m^3 in the top 1.5 m over the course of the study. The regression model of the EM38 vertical dipole mode data with water content to the 1.5 m depth for all locations monitored on the capillary barrier design resulted in a standard deviation of 0.016 m^3/m^3. Horizontal dipole mode data correlated with water content to the 0.75 m depth had a standard deviation of 0.022 m^3/m^3 for all locations on both barrier designs. Models at individual survey locations generally exhibited much smaller standard deviations, ranging from 0.005 to 0.018 m^3/m^3 and averaging 0.010 m^3/m^3. The smaller standard deviations and general similarity of regression slope values of the models at individual locations indicate that this technique is more accurate as an indicator of changes in water content than as an indicator of the absolute value of water content at a given location. Sources of variability were attributed to horizontal and vertical variation in soil salinity, the vertical distribution of water at the time of a particular survey, and subtle differences in topsoil thickness and surface roughness. Results indicate that electromagnetic induction is useful for evaluating infiltration. The EM technique resulted in standard deviation values for water content similar to those of the neutron probe method but is capable of monitoring larger areas much more rapidly and at a lower cost.