Simulation of close range remote sensing of subsurface features using GPR for urban utility information system development

摘要

In rapidly urbanizing old cities underground utility maps or urban utility information systems seldom exist. Close range remote sensing using Ground penetrating radar (GPR) for detection of the buried utilities is increasingly becoming common. GPR can help in developing or supplementing and updating an urban utility information system. The GPR data (radargram) needs to be interpreted by applying knowledge of buried utility responses under the influence of (a) antenna center frequency (b) host medium relative permittivity, conductivity and (c) object shape, size and material. In view of impracticality of generating utility response information directly in the field, Finite difference time domain (FDTD) simulation of radar wave propagation is carried out. The present work describes the database generation of GPR responses through simulation using an exclusive software GprMaxV2.0. A buried utility pipe produces a hyperbolic pattern in the radargram. In general, utilities may be made of metal, concrete or PVC; may lie within a shallow depth of about 0.5m-1.0m and their diameters may range upto to 1.0m; the relative permittivity of a dry soil could vary from about 4 to 15. Considering these aspects, problem of a pipe buried in soil is formulated, radargrams are simulated and variations of amplitudes and hyperbolic patterns are studied. To minimize the time-intensive simulations, Response surface method (RSM) is used to model amplitudes and hyperbolic patterns as functions of their influencing parameters. A database of simulated responses along with RSM modeling is seen to be a useful component of or complement to an urban utility information system.