Inversion Strategy - Models and Methods of Magnetotellurics - page 525

Fig. 12.62 Algorithm of partial inversions

the observed values of the real and imaginary tippers. The TP model seems to agree
well with observations: at most stations, the misfits

Im W zy are at
least 5-10 times less than the maximum tipper variations. A noteworthy feature of
the model is a conductive continental asthenosphere with a branching-out vertical
low-resistivity zone penetrating the continental crust under the High Cascades. This
feature of the TP model distinguishes it from EMSLAB-I and EMSLAB-II mod-
els and makes it similar to the predictive CASCADIA model, in which a vertical
high-temperature zone of wet and dry melting characterized by low resistivities is
localized beneath the High Cascades. When eliminating the conductive continen-
tal asthenosphere and vertical low-resistive crust-mantle zone, we get the misfits
increased by factors of 1.5-2.5. So, we conclude that the tipper inversion settles the
dispute among the EMSLAB-I, EMSLAB-II, and CASCADIA model in favor of
the CASCADIA model.
Level II . Inversion of

Re W zy and

(TE-mode). On this level, we check and edit the results
of tipper inversion. The inversion of the longitudinal

-curves distorted by near-
surface heterogeneities requires a preliminary normalization of the apparent resis-
tivities, which almost always involves the risk of errors (especially in mountains).
Therefore, interpreting the EMSLAB data, we limit ourselves to the inversion of the