Accounting for Modeling Errors in Linear Inversion of Crosshole Ground?Penetrating Radar Amplitude Data: Detecting Sand in Clayey Till

Mapping high permeability sand occurrences in clayey till is fundamental for protecting the underlying drinking water resources. Crosshole ground penetrating radar (GPR) amplitude data have potential to differentiate between and clay, can provide 2D subsurface models with a decimeter-scale resolution. We develop probabilistic straight-ray-based inversion scheme, where we account forward modeling error arising from choosing straight-ray solver. The described by Gaussian probability distribution included total noise model addition of covariance models. Due linear formulation, are able decouple traveltime obtain results fast. evaluate approach through synthetic study, inverted slowness attenuation tomograms using several scenarios. find that accounting successfully without artifacts. This especially case since structure significantly different other Overall, field confirms study; however, performs slightly better than inversion. characterize 0.4–0.6 m thick layer as well internal variations matching observed geological information borehole logs excavation.