Adding realistic noise models to synthetic ground‐penetrating radar data

Abstract Cost‐effective computing capabilities paved the road to use numerical modeling develop advanced methods and applications of ground‐penetrating radar (GPR). Realistic synthetic data corresponding techniques, respectively, should consider all subsurface above‐ground aspects that influence GPR wave propagation characteristics recorded signals. Critical can be realized in modern tools include heterogeneous frequency‐dependent material properties, complex structures interface geometries as well 3D antenna models, including interaction between subsurface. However, realistic noise related electronic components a system or ambient electromagnetic is often not considered simplified by assuming white Gaussian model which added modeled data. We present an approach scenarios typically observed field into flow In our approach, we extract from traces add it via convolution‐based process. illustrate methodology using exercise, where contaminate 2D dataset with urban environment. Comparing noise‐contaminated similar environment, illustrates method allows generate further highlights limitations pure models. This article protected copyright. All rights reserved