Accounting for modeling errors in linear inversion of crosshole ground-penetrating radar amplitude data: detecting sand in clayey till

Abstract

Heterogeneous glacial sediments, such as clayey till, dominate large parts of the near-surface geology of the Northern Hemisphere (Houmark-Nielsen, 2010). Sand layers and lenses control water and contaminant flow pathways in the otherwise low-permeable clay matrix. Delineation and characterization of these sand structures and bodies are necessary to determine the timing, the amount and the quality of the water percolating through these sediments (e.g., Gravesen et al., 2014).A method for mapping these sand occurrences is by using crosshole ground penetrating radar (GPR). Crosshole GPR is a fast, minimally invasive, electromagnetic (EM) method, which is based on transmission of radio frequency EM waves, traveling from a transmitter, located in one borehole, to a receiver located in a neighboring borehole. The recorded traveltime and amplitude of the wave provide information on subsurface dielectric properties, which can be linked to parameters important for flow and transport processes, such as volumetric