A COMBINED SEISMIC REFLECTION AND REFRACTION STUDY OF A LANDFILL AND ITS HOST SEDIMENTS

Abstract

In an attempt to delineate the base of a landfill and map the geometries of the host sediments, we have recorded a high-resolution seismic profile. To obtain sufficient resolution in the heterogeneous landfill environment, common midpoint (CMP) spacing was set to 0.125 m and subsurface coverage (i.e. fold) was maintained at ≥120 in the central region of the survey. Despite the high density and high redundancy of the data, severe source-generated noise (i.e. direct, refracted, guided and surface waves) and strong lateral velocity variations made it difficult to identify reflections on processed shot and CMP gathers. However, a quasi-continuous sequence of reflections R1–R3 was eventually traced along the length of the profile. After time-to-depth converting the stacked seismic reflection section using poorly resolved initial stacking velocities, no consistent correlations with boundaries identified in nearby boreholes and on three-dimensional georadar data were apparent. In a first attempt to obtain more reliable velocities, ∼183,000 first-arrival times were tomographically inverted. Unfortunately, the resultant velocity model was found to be incompatible with knowledge supplied by the borehole and georadar data and the seismic reflection section. By including the known depths to a key geological horizon and the R1–R3 traveltimes as constraints, a second suite of tomographic inversions produced a satisfactory model. This model included a thin capping layer of humus and sandy clay (velocities of 400–1000 m/s) overlying a distinctly lower velocity landfill (200–600 m/s) along the northern half of the profile and a southward thickening sequence of fluvial deposits (600–900 m/s) along the southern half. A southward thinning layer of compact lacustrine sediments and basal till (2000–3800 m/s) and a nearly horizontal bedrock interface (4000–5400 m/s) was mapped beneath the entire profile. Although independent applications of the seismic reflection and refraction techniques were not successful in meeting the survey objectives, a combination of the two approaches suitably constrained by borehole information finally provided the required details on the landfill and surrounding sediments. Nevertheless, our study has highlighted the limitations of employing 2-D seismic refraction and reflection methods for resolving problems in highly heterogeneous 3-D media