Repeated Electromagnetic Induction Surveys for Understanding Landscape Soil and Water Dynamics.

Apparent soil electrical conductivity (ECa) measured by electromagnetic induction (EMI) techniques has been widely used to interpret soil spatial variability. However, limited studies have been done to repeat such measurements over time in the same location to detect temporal changes in soil and water processes. We report here the merits of repeated EMI surveys that can be used more extensively to reveal temporal dynamics in heterogeneous soilscapes, especially those related to soil water. Repeated EMI surveys were conducted from 1997 to 2009 over a 19.5-ha agricultural field located in central Pennsylvania. The first two surveys were done in 1997 and 2006 and compared different meters (EM38, EM31, and Dualem-2), dipole orientations, and geometries. Another six surveys were conducted in different seasons in 2008-2009 using the same EM38 operated in vertical dipole orientation (EM38V). The ECa collected during the wetter periods (> 10 mm rain during the previous seven days) showed significantly greater spatial variability (greater sills and shorter spatial correlation lengths), indicating the influence of soil water redistribution on soil ECa. Over a relatively short time period (e.g., a few months as used in this study), most soil properties controlling ECa (e.g., texture and depth to bedrock) can be assumed to remain unchanged, except soil moisture. Thus repeated EMI surveys would reflect soil water temporal dynamics and related flow paths in the landscape. Higher ECa values were found in areas close to simulated subsurface flow paths, especially during the wetter periods. In areas wetter than the overall average of the study area, correlation between the relative difference in ECa and in soil moisture was high (R² = 0.59-0.77), but not in dry areas. The measurement depth of EMI determined the efficacy of capturing soil spatial variability, therefore the EM31 operated in vertical dipole orientation (EM31V) with the deepest measurement depth (6 m) had the closest relation to depth to bedrock in the study area (R² = 0.58). The EM31 operated in horizontal dipole orientation (EM31H) and Dualem-2 operated in horizontal co-planar geometry (EMHCP), with measurement depth of 3 m, showed the best correlation with silt content (R² = 0.45-0.47) because the top 2 m of the soil profiles exhibited distinct difference in silt content. Since the soil series in the study area differ mainly in depth to bedrock, silt content, and soil moisture, mapping soil series in this landscape using EMI should be limited to proper measurement depths and wet periods in order to obtain the best possible results. Overall, the optimal use of EMI will depend on soil properties targeted, the landscape situation, the specific EMI meter and its operating setting, as well as the timing of survey. Repeated EMI surveys add valuable information about the temporal changes of spatially-distributed soil and water properties and the related underlying processes.