We address the inverse problem to recover earth parameters from Electrical Resistivity (ER) and Ground Penetrating Radar (GPR) data measured at the surface of the earth. Robust least squares inverse methods are commonly used in practice. However, they result in smooth parameter estimates, which is not desirable in many geological and hydrogeological applications because it is common to encounter sharp boundaries in the subsurface.
It has been shown that when a least squares smoothness constraint is relaxed at a known boundary, sharp discontinuities can be recovered. Here we use GPR data to constrain the boundary in least squares inversion of ER data. We also develop a methodology to determine when a boundary estimate is a good choice relative to other boundary estimates. Results from both synthetic and field data from the Boise Hydrogeological Research site will be shown.