A new vertical continuation procedure for airborne electromagnetic field data from the modified image method

Abstract

Airborne electromagnetic (AEM) data measured by equipment in a bird tethered to a helicopter have large variations caused by the unavoidable vertical excursions of the helicopter as it traverses its flight path. Such large changes tend to mask the smaller changes in field strength caused by lateral variations in the earth’s electrical conductivity along the flight path, which is the information that is the goal of AEM surveys. Signals produced by conductivity anomalies such as sea‐ice keels and pipelines in marshes or in the shallow ocean are enhanced and may be apparent directly in the continued fields. Furthermore, electronic or environmental noise is more easily detected in the continued fields and reduced by various methods of filtering and signal processing. In the modified image method (MIM) formalism for AEM fields, the algebraic expression for the secondary to primary field ratio [Formula: see text] is given in terms of R, where R is the total complex vertical separation of the primary and image dipoles [Formula: see text] scaled to the coil spacing ρ, [Formula: see text] is the complex effective skin depth, and h is the altitude of the bird. An inverse algebraic relation gives R as a function of [Formula: see text]. In this paper we present a simple and accurate method of continuing the field by way of continuing R. Because R is linear in h, the vertical continuation of R from h to h0is accomplished by a simple linear translation. This method is applied to a flight line of an AEM survey of Barataria Bay, Louisiana, which includes both the marsh and a near‐shore region of the Gulf of Mexico. The smoothnesss of the continued data over the Gulf implies that the variability of the continued data over the marsh is attributable to horizontal variation in salinity, soil porosity, and water depth rather than noise. To produce more accurate values for R, we have also included details of an extended half‐space renormalization function which, in effect, removes residual differences between the fields calculated from the MIM algebraic and the numerical evaluation of the exact Sommerfeld integral representations of the [Formula: see text] field.