Introduction
To a large part, the Magnetotelluric project work is geologically or interdisciplinary oriented but our research also requires methodological developments. Many of the novel approaches discussed below are originally academic theses which were compiled at the GFZ and in cooperation with universities.
Data processing
For MT measurements, orthogonal components of the electric and magnetic fields are recorded as time-variations at the surface of the earth. In the frequency-domain representation, the electric and horizontal magnetic field components are linearly related by the impedance tensor Z, a so-called response function of the earth. The aim of electromagnetic data processing procedures is to estimate these response functions as accurately as possible.
Crucial for the outcome of any data processing algorithm is the amount of noise on the horizontal components of the magnetic field which can vary considerably in time, between sites and over the frequency range. The removal of such contaminated data beforehand is essential for most data processing schemes, as the magnetic channels are usually assumed to be free of noise. With the method described in Ritter et al., 1998 source field irregularities are removed from the data, thereby providing suitable working conditions for the robust data processing procedures.
Robust remote reference processing can improve the data quality at a local site, but only if synchronous recordings of at least one additional site are available and if electromagnetic noise between these sites is uncorrelated. If these prerequisites are not met, Weckmann et al., 2005 suggest an alternative approach for noise removal, based on a combination of frequency domain editing with subsequent single site robust processing.
Oettinger et al., 2001 discuss a method for noise reduction in MT time series with a new signal-noise separation method and Müller et al., 2000 introduce a new method to compensate for bias in magnetotellurics.
Tensor decomposition and imaging methods
Weckmann et al., 2003a present a method for converting the magnetotelluric (MT) impedance tensor into an apparent resistivity tensor. Contrary to other methods, the results of this new method (called PNA) are stable and significant under extreme 3-D conditions. Application of PNA to MT data from the Waterberg Fault in Namibia unravels a complicated 3-D impedance and reveals a clear correlation between the resistivity tensor and the surface geology.
Imaging of seismic scatterers and magnetotelluric soundings reveal a sharp lithologic contrast along a 10 km long segment of the Arava Fault (AF), a prominent fault of the southern Dead Sea Transform (DST) in the Middle East (Maercklin et al., 2005).
Data interpretation
A large share of our published work focuses on the interpretation of electrical conductivity models in view of the geological or geodynamic setting. Hoffmann-Rothe et al., 2004 examine the relationship of conductivity and internal architecture of a fault zone by comparing the record of structural deformation across a fault zone with its electrical conductivity image. A comparison of four major fault systems shows both similarities and marked differences in their electrical subsurface structure. Differences in the electrical structure of these faults within the upper crust may be linked to the degree of deformation localization within the fault zone (Ritter et al., 2005).
The role and detection of electrical anisotropy in the subbsurface of the earth is still a highly disputed topic. In a magnetotelluric study of the Damara Belt in Namibia, MT phases over 90° reveal the internal structure of the Waterberg Fault/Omaruru Lineament as a highly anisotropic zone (Weckmann et al., 2003b). Eisel and Haak, 1999 discuss the existence of macro-anisotropy in the electrical conductivity in the crust of the German Continental Deep Drilling site (KTB).
Magnetotelluric and seismic methods provide complementary information about the resistivity and velocity structure of the subsurface on similar scales and resolutions. Independently derived inverse models from these methods can be combined, using a classification approach, to map geologic structure (Bedrosian et al., 2007).
Becken et al. (2006) analyse properties of the TE and TM modes in a 3D environment. They show that TM mode galvanic excess currents of shallow origin can overprint the electric field response of deep structures. Since the MT impedance tensor is a mixed-mode transfer function, the TM mode response of shallow origin does not only dominate over the TM mode response from deep structures, but also distorts the TE mode parts. Hence, the MT impedance tensor does not always reflect the properties of the deep earth, if the shallow earth is heterogenous.
Becken and Pedersen (2007, in prep.) developed a technique to inter- and extrapolate measured magnetic transfer functions in a self-consistent way. The analysis is based on an equivalent model and allows the prediction of the magnetic fields and transfer functions anywhere within and outside of a MT array. The method can also be applied as a consistency check of observed data.
Other methodological developments were by Schwalenberg et al., 2002 who applied sensitivity studies to a 2-D resistivity model from the Central Andes and by Lezaeta and Haak, 2003 who examine strong current channeling with magnetotelluric phases over 90°.