抄録
Diverse magnetotelluric data (apparent resistivity, phase or geomagnetic transfer function) used during the inversion process provide different information on the model as a consequence of data error. The relative influence of these data constitutes a subject of interest on the inversion process. This influence can be evaluated from the error ratio between two types of data; thus, when phase and the logarithm of apparent resistivity are involved, the wellknown ratio of one half is obtained. A new error ratio between the geomagnetic transfer function and the logarithm of apparent resistivity is presented. We deduced this ratio, which is bounded by one half of the amplitude of the geomagnetic transfer function. In order to verify this new ratio, we employed a technique based on the study of the RMS misfit, obtained after an intensive inversion computation whilst taking different error values for the different data. This technique was applied to synthetic and experimental data, and the results agree with the proposed value. This value should be taken into account for setting error floors when performing joint data inversion in order to obtain the same influence from the different data.
