Magnetotelluric studies from two contrasting Brazilian basins: a reassessment of old data

Before the mid-1980s large numbers of magnetotelluric (MT) data were collected and interpreted without addressing the possible effects of near-surface (e.g. static) distortion contributions. In the intervening time, modelling and interpretation techniques have improved. There are many instances where extensive, processed tensor survey data of reasonable quality exist on various back-up media. Before any complete reanalysis of these often large data sets, it is useful to investigate the degree of the distortion effects and the approximate degree of resolution of the data. Such assessment should help to clarify whether a full tensorial reanalysis is worthwhile. The present study provides contrasting examples of the reappraisal of old data sets collected in two Brazilian basins, both with associated seismic reflection profiles and deep exploration well-logs. Both examples use 'simplified' data consisting of only the off-diagonal tensor elements. These data are used in a limited way to assess static distortion and their intrinsic resolution capabilities. In the first example, from the basalt-covered Parana basin, a very high degree of static distortion is displayed. The application of simple distortion correction techniques allow the data to define the main (vertical) geoelectric units of the basin. In the second example, from the Solimoes basin in northern Brazil, static distortion, although less severe, can be readily identified. In this complex petroleum province the resolution of structure, both vertical and lateral, is of very low order. The MT data appear 'blind' to the detail of the Palaeozoic sedimentary rocks, and more detailed reanalysis is probably not warranted.