Focused attributes derived from AEM surveys using the continuous wavelet transform

Abstract

Interpretation of a hydrogeophysical survey is a complex and comprehensive process. In addition to an areal coverage with AEM data, most often an interpretation involves additional data that are time consuming to collect and complicated to integrate into an overall model, e.g. borehole logs, borehole core samples, water chemistry, surface vegetation, satellite imagery plus the generally accepted geological background knowledge. Compared with the complexities of the interpretation process, the acquisition, QC and inversion of AEM survey data are a more straightforward affair and considerably less time consuming. Interpretation basically has to do with identifying categories and finding boundaries between them so that depths, thicknesses, lithologies and a whole range of other model attributes can be estimated, qualitatively and quantitatively. To supplement the traditional product delivered by the inverter to the interpreter: inversion models displaying the distribution of subsurface electrical conductivity, I present two methods based on the Continuous Wavelet Transform that can deliver more focused attributes to assist in the interpretation. In the first method, layer boundaries in the smooth multi-layer models that are most often used in the inversion of large data sets are found. In the second method, the spatial distribution of the natural categories of the model parameter is found. Both methods are based on the inversion models and, evidently, they are useful to the extent that the variations in conductivity reflect geological / hydrogeological boundaries and categories - which is for the interpreter to decide.