Noviembre 22, 2016 12:00 pm CST, Sala de Conferencias del SisVOc
Philip Hering, Andreas Junge
Institute of Geosciences, Applied Geophysics, Goethe University Frankfurt am Main, Germany
During the evaluation of magnetotelluric (MT) data, frequency-dependent, complex transfer functions between magnetic (B) and electric (E) fields are calculated. These transfer functions provide information about the conductivity of the subsurface and thus are of crucial importance. Nevertheless, in some cases the data processing proves to be difficult, since the recorded timeseries can be heavily contaminated by anthropogenic noise signals, e. g. galvanic currents or near field sources. Several common methods, like robust or remote reference processing, address these problems, however in case of coherent noise sources they might fail (Junge, 1996). As a consequence the differentiation between coherent noise and MT-signal is still a big challenge in magnetotellurics and will therefore occupy a key role in this talk.
Initially, the talk will give a short summary about the principles of the magnetotelluric method. This includes e. g. data acquisition, calculation of transfer-functions and conclusions on geology. Subsequently, the focus will be laid on robust estimation procedures of transfer-functions. Especially a multivariate processing scheme (Hering, 2015), based on singular value decomposition, will be discussed in more detail. Hereby, effects of coherent noise sources can be identified and minimized by applying an eigenvalue selection criterion. Application of the method will be shown within an example from the Odenwald, a low mountain range located close to the strongly urbanized Rhein-Main area in central Germany.
Archivos adjuntos: Conferencia Magistral - Philip Hering