Premium
Evaluation of source signature during Full Waveform Inversion for seismic reconnaissance in tunneling
Author(s) -
Riedel Christopher,
Musayev Khayal,
Baitsch Matthias,
Hackl Klaus
Publication year - 2019
Publication title -
pamm
Language(s) - English
Resource type - Journals
ISSN - 1617-7061
DOI - 10.1002/pamm.201900089
Subject(s) - inversion (geology) , seismogram , geology , signature (topology) , communication source , computer science , frequency domain , waveform , acoustics , seismology , geometry , physics , telecommunications , mathematics , radar , computer vision , tectonics
Abstract By transmitting seismic waves in the ground the spread, reflected and refracted waves can be recorded at different positions. The arising seismograms contain information of the geology, which can be evaluated with the Full Waveform Inversion (FWI). For the application of FWI for reconnaissance in tunneling extensive information has to be provided to achieve an appropriate model of the geology in front of the tunnel face. Such information are the position and geometry of reflecting boundaries, exact position of sources and receivers, a proper initial model of the distribution of the material parameters. The signal, which is transmitted by the senders, has also an influence on the inversion. Even if the intended source function is known, the effective signal depends on the transfer functions between sender‐soil and soil‐receiver and its identification is challenging. To avoid cumbersome investigations on the source function in advance we applied a simultaneous identification of the source signature during the FWI in the frequency domain. For synthetic models of a 2D tunnel configuration with disturbances, we will show how the calculated source signature of a frequency converges to the real source signature during the iterations of the FWI. The influence of different sender and receiver configurations will be analyzed. Furthermore, the influence of the simultaneous evaluation of the source signature on the results of the FWI will be evaluated. In a post‐processing step of the FWI, the source signature of the whole frequency spectrum will be determined.