Python based modelling of ground penetrating radar data

Nowadays, non-destructive identification of buried structures in archaeological sites has become important. Ground penetrating radar method (GPR), one of the shallow geophysical methods, is one of the most effective methods in revealing sarcophagi, walls, roads and similar structures. Especially before the excavation works begin, mapping the area, narrowing the search area and locating existing objects quickly and economically are very important in terms of budget and time. In addition, it is also used in areas such as evaluation of critical infrastructures such as bridges and roads, mapping of glaciers, detection of anti-personnel land mines, geological-geotechnical research, mining research, and engineering structures research. In this study, underground models were produced using the finite difference method in the time domain with the Python-based gprMax software. In the first phase of the work, structures with different physical properties were modeled, and then an area of 200 square meters was determined in the Yassitepe Mound located in Bornova district of Izmir Province. In the area determined as the study area, measurements were made in the form of two grids of 10x10 meters and at 0.5 meter intervals. Field studies were carried out with a Proceq GS8000 model ground radar that collects measurements in the 40-3440 MHz modulated frequency range. As a result of the measurements, 2D level amplitude maps, 3D images were created, and wall structures were determined. The theoretically produced models were compared with the data collected from the field, and the excavation sites were suggested.