Reconstruction of ground penetrating radar backscatter of subsurface features for utility mapping

Abstract

Ground Penetrating Radar (GPR) is one of the trenchless technologies; widely used for subsurface utility detection and mapping. However, in the context of subsurface utility mapping, constraints of achieving specific accuracy requirements so far have not been addressed by each system manufacturer and users. This research investigates the utilization of GPR for subsurface utility mapping, with the following specific objectives: (i) to design and built a calibration site for analysing locational and detectability accuracies of GPR; (ii) to examine and analyse the effects of GPR data acquisition approaches to the locational and detectability accuracies; (iii) to characterize GPR backscatter for recognition of utility’s material based on digital image processing techniques for retrieving the uniqueness of backscatters from respective utilities, and (iv) to examine and model GPR backscatters constraints for detecting and mapping stacked subsurface utilities in both vertical and horizontal orientations. Dual frequencies (250 and 700 MHz) GPR system was used in this study, experimented in both lab controlled and in-situ environments with settings of the system and scene parameters. Optimum values obtained in the lab for both system and scene parameters were then adopted for acquisition of data from in-situ measurement and also used in the Finite-Difference Time-Domain (FDTD) numerical modelling for validating the results of the study. Results of this study contributed three main findings: (i) the GPR locational and detectability accuracies for subsurface utility mapping are directly proportional to the data acquisition scanning techniques, where the ‘along-pipe’ scanning, which is rarely practised in the industry, yields the best locational and detectability accuracies, confirming to Quality Level A utility data; (ii) GPR backscatters with appropriate treatment can yield unique backscatter signature for recognition of utility’s material, hence, opening a platform for new valuable addition to the GPR application for utility’s material recognition besides utility detection and localization of buried utility; and (iii) the locational and detectability error trend and constraints of GPR measurements within crowded subsurface utility infrastructures yield a “best practice” procedure for determining the safe buffer zone for maintenance works; very crucial aspects in installation of new utility infrastructure and detecting aging utility.