Common-reflection-surface imaging of shallow and ultrashallow reflectors

Abstract

We analyzed the feasibility of the common-reflection-surface (CRS) stack for near-surface surveys as an alternative to the conventional common midpoint (CMP) stacking procedure. The data-driven, less user-interactive CRS method could be more cost efficient for shallow surveys, where the high sensitivity to velocity analysis makes data processing a critical step. We compared the results for two field data sets collected to image shallow and ultrashallow reflectors: an example of shallow P-wave reflection for targets in the first few hundred meters, and an example of SH-wave reflection for targets in the first 10 m. By processing the shallow P-wave records using the CMP method, we imaged several nearly horizontal reflectors with onsets from 60 to about 250 ms. The CRS stack produced a stacked section more suited for a subsurface interpretation, without any preliminary formal and time-consuming velocity analysis, because the imaged reflectors possessed greater coherency and lateral continuity. With CMP processing of the SH-wave records, we imaged a dipping bedrock interface below four horizontal reflectors in unconsolidated, very low velocity sediments. The vertical and lateral resolution was very high, despite the very shallow depth: the image showed the pinchout of two layers at less than 10 m depth. The numerous traces used by the CRS stack improved the continuity of the shallowest reflector, but the deepest overburden reflectors appear unresolved, with not well-imaged pinchouts. Using the kinematic wavefield attributes determined for each stacking operation, we retrieved velocity fields fitting the stacking velocities we had estimated in the CMP processing. The use of CRS stack could be a significant step ahead to increase the acceptance of the seismic reflection method as a routine investigation method in shallow and ultrashallow seismics.