COUPLED INVERSION OF ELECTRICAL RESISTIVITY AND HYDROLOGICAL MODELS TO QUANTIFY SOIL MOISTURE DYNAMICS BELOW A MICHIGAN ECOTONE

Anticipating how changes in land use and climate will impact water budgets requires understanding the relationship between soil moisture and vegetation across space and time. Unfortunately, traditional geophysical methods for capturing field-scale processes of this nature, such as remote sensing and in situ point measurements, have their limitations. Electrical resistivity (ER) however, has been identified as a potentially powerful geophysical tool for studying root-zone moisture dynamics at sub watershed scales. Despite recent advances, interpretation of geophysical data often occurs independently from hydrological observations, causing uncertainty and propagating measurement errors. This research uses 1D ER soundings with a coupled hydrogeophysical inversion model to determine the impact of changing vegetation types on soil moisture. The goal is to develop a more universal method for improving predictions of hydrological properties at field-scales through the use of coupled inversion techniques.

The Kellogg Biological Station near Battle Creek, MI, provides a unique setting for this study, with a shift from mature forest to young forest, shrub, and finally grass. Across this ecotone, graphite electrodes have been permanently installed at 1.5m intervals along a 166.5m transect, with additional electrodes for the 1D soundings. Six sounding locations, roughly corresponding to each vegetation type, allow for shallow ER measurements. Reciprocal ER measurements with a-spacings of .5, .75, 1.5, 3, 4.5, 6, 7.5, 10.5, and 13.5m were conducted at each sounding location in June, August, and October 2012. Deeper measurements with a-spacings of 18, 24, 36, and 54m were collected in August and October only. Early interpretations suggest that seasonal and vegetative differences impact soil moisture distribution and plant water use. The coupled inversion model integrates four components. A hydrological model takes in parameters such as temperature, precipitation, soil and vegetation type, and outputs water content, which a petrophysical model converts to obtain 1D resistivity profiles. A forward geophysical model uses those values to determine apparent resistivities for comparison to those measured in the field. The outcome is then used for optimizing the hydrological model parameters.