The effects of receiver arrangement on velocity analysis with multi‐concurrent receiver GPR data

Abstract

Determining subsurface electromagnetic (EM) wave velocity is critical for Ground-penetrating radar (GPR) data analysis, as velocity is used for the time-to-depth conversion, and hence leads to obtaining the precise location of the objects of interest. Currently, the way to acquire detailed subsurface EM wave velocity models involves employing multi-offset GPR surveys, such as wideangle reflection-refraction (WARR), in conjunction with normal moveout (NMO) based velocity analysis. Traditionally, these surveys are carried out using two separate transducers and were therefore time-consuming and had limited uptake. Recent advances in GPR hardware have allowed the development of novel systems with multi-concurrent sampling receivers, which enable rapid and dense acquisition of WARR data. These additional receivers increase the overall size, weight, and cost of the system. Therefore, we investigated the effects of receiver arrangement on NMO-based velocity analysis, and considered reducing the overall number of transducers, whilst maintaining satisfactory velocity spectra resolution, and hence, obtaining detailed stacking velocity models as well as improved stacked reflection sections. We used both simulated data from complex threedimensional (3D) models as well as field data and examined different numbers and positions of receivers in different environments. Our results show that velocity spectra resolution can be maintained within acceptable limits whilst reducing the number of receivers from a configuration with seven equally spaced receivers, to a sparse configuration of four receivers. Thus, being able to decrease the number of receivers used by these new GPR systems will reduce both the total system weight and cost, and hopefully, increase their adoption for GPR surveys.

Finetuning ground penetrating radar velocity analysis from hyperbola fitting using migration

“…Individual layer velocities can be found using Dix formulae (Dix, 1955). Angelis et al (2022) describe how a multi-concurrent sampling receiver tool, the 'WARR machine' , can be used to fast collect WARR data, and how NMO corrections are used to find an optimum receiver configuration to obtain satisfactory velocity spectra resolution. However, this method is not suited for a deep investigation like the one at Longyearbreen at Svalbard.…”

GPR velocity correction method in transversely heterogeneous media based on CMP data

“…In recent years, methods for velocity inversion based on CO data have been developed, which are suitable for areas where it is difficult to acquire multi‐offset data (Forte et al., 2014; Leong & Zhu, 2021). WARR data is typically combined with normal movement to achieve velocity estimation (Angelis et al., 2022; Kaufmann et al., 2020). As a kind of WARR data, CMP data has been widely used in velocity estimation.…”