2-D joint inversion of semi-airborne CSEM and LOTEM data in eastern Thuringia, Germany

Abstract

SUMMARY Various electromagnetic (EM) techniques have been developed for exploring natural resources. The novel frequency-domain semi-airborne controlled source electromagnetic (semi-AEM) method takes advantages of both ground and airborne techniques. It combines ground-based high-power electrical dipole sources with large-scale and spatially densely covered magnetic fields measured via airborne receivers. The method can survey the subsurface down to approximately 1000 m and is particularly sensitive towards conductive bodies (e.g. mineralized bodies) in a more resistive host environment. However, the signal-to-noise ratio of semi-AEM is lower than that of ground-based methods such as long-offset transient electromagnetics (LOTEM), mainly due to the limited stacking time and motion-induced noise. As a result, the semi-AEM often has reduced depth of investigation in comparison to LOTEM. One solution to overcome these flaws is to analyse and interpret semi-AEM data together with information from other EM methods using a joint inversion. Since our study shows that LOTEM and semi-AEM data have complementary subsurface resolution capabilities, we present a 2-D joint inversion algorithm to simultaneously interpret frequency-domain semi-AEM data and transient electric fields using extended dipole sources. The algorithm has been applied to the field data acquired in a former mining area in eastern Thuringia, Germany. The 2-D joint inversion combines the complementary information and provides a meaningful 2-D resistivity model. Nevertheless, obvious discrepancies appear between the individual and joint inversion results. Consequent synthetic modelling studies illustrate that the discrepancies occur because of (i) differences in lateral and depth resolution between the semi-AEM and LOTEM data caused by different measuring configurations, (ii) different measured EM components and (iii) differences in the error weighting of the individual data sets. Additionally, our synthetic study suggests that more flexible land-based configurations with sparse receiver locations are possible in combination with semi-AEM without a significant loss of target resolution, which is promising for accelerating data acquisition and for survey planning and logistics, particularly when measuring in inaccessible areas.