Transient electromagnetic soundings in complex geological environments.

TEM sounding curves are distorted by geological noise in complex environments and often cannot be interpreted using practical 1D interpretation schemes which assume homogeneous, horizontal strata. Systematic experiments with a Geonics EM42 and MarkI and MarkII SIROTEM systems in complex environments show that geological noise is dominant over system noise and errors in loop receiver geometry which seldom exceed 2% of the measured signal. Inversion results in complex areas depend on the loop configuration and type of receiver, unlike those from 1D situations. Small loops are less susceptible to geological noise and have a better vertical resolution than large loops which, however, do have greater depths of penetration. When carrying out soundings along profiles it is important to match the station spacing and loop size to the scale of lateral discontinuities. Data from four geologically complex areas in Cyprus, Spain, Northern Ireland and England have been collected and interpreted to investigate the applicability of using TEM depth soundings in such environments. The results have shown than in some areas, soundings can provide useful quantitative information on depths and resistivities. Even in extremely noisy areas the method can be used to locate anomalies for follow up work with other geophysical methods. Investigations of various interpretation schemes including pseudosections, time slices and 1D inversions have demonstrated limitations in complex environments. A new interpretation scheme based on a direct calculation of approximate depths from an apparent resistivity curve has been developed which removes some of the ambiguities due to the resistivity dependent velocity of the EM waves. A general purpose TEM reduction and interpretation program has been written to carry out apparent resistivity and direct depth calculations, forward modelling and 1D inversions. Tests on synthetic and field data show a good agreement with sections produced from inversions and borehole data.