Broadband constrained inversion of common offset GPR data using prior information from well log

【摘要】：Quantitative amplitude information of ground penetrating radar(GPR) spawns another generation of data analysis and interpretation tools based on inversion method to image subsurface material properties.For the frequently used common offset GPR surveys,impedance inversion provides an effective way for estimating parameters of interest.However,previous studies are limited to band-limited impedance inversion,where errors would accumulate with computing the impedance.Using prior information from well log to construct a dependable starting model,we present a novel scheme inverting common offset GPR amplitudes to subsurface properties by means of broadband constrained inversion.After introducing the theoretical background,we show a numerical test which employs a realistic and complex stochastic model.The broadband constrained inversion result corresponds well with true model while it also exhibits advantages over band-limited impedance inversion result on accuracy.It is shown that broadband constrained inversion of common offset GPR data is a reliable method in near surface geophysics investigations.