GNGTS 2013 - Atti del 32° Convegno Nazionale

Accurate evaluation of edges and dip of faults and contacts through the Volume Upward Continuation (VUC) of gravity data D. Mastellone, M. Fedi, V. Paoletti DISTAR, Department of Earth Sciences, Environment and Resources, University of Naples Federico II, Italy Introduction. Several methods have been used to retrieve values of dip of faults by using lateral offsets of the zero-crossover point of the second horizontal derivative of upward- continued gravity profiles. The use of derivatives of potential fields represents a signal enhancement technique experimented since a long time. It allows the information content of the signal to be enhanced without implying any arbitrary assumption. Often, such methods are coupled with upward continuation, which is used to transform anomalies measured on one surface into those that would have been measured on some higher altitude surface. In this paper we demonstrate the efficiency of a new approach to upward continue potential field data, the Volume Upward Continuation (VUC), coupled with traditional horizontal derivative techniques. By using VUC it is immediate showing that upward continuation of the signal to higher altitudes yield information about progressively deeper sectors of the discontinuity. In particular, looking at the position of the maxima of the horizontal derivative of the continued field, we can observe that they will be laterally shifted toward the dipping direction of the discontinuity, in a way proportional to the continuation height. We applied VUC followed by horizontal derivative to a gravity profile extracted from a gravity survey on the Venelin-Aksakov fault in Bulgaria, in order to get further information on the dipping direction of this structure. Upward continuation and horizontal derivatives. Upward continuation operator uses measurements of a field at one elevation, level or surface to determine the values of the field at a higher level. Upward continued data can be calculated by convolution in either the space domain or the Fourier domain. In this last domain the Fourier transform of the data is simply multiplied by the frequency filter: (1) where k is the wavenumber vector and Δz is the distance between the original surface and the final one. Real data are discrete and refer to a finite survey area; so, when using circular convolution to calculate upward continuation in the frequency domain, aliasing errors can affect the low frequency content of upward continued data. These errors can be reduced by performing the Fourier transform on a larger dataset, which spreads outside the survey area (Oppenheim and Schafer, 1975; Fedi et al. , 2012), built with other surveys data or, through extrapolation algorithms (zero-padding, maximum entropy extension, symmetric extension). Upward continuation is mostly helpful to enhance the effects of deep sources, as it attenuates the highest frequency content of the signal, which is usually associated to shallow sources; or to trace the dipping direction of oblique discontinuities (e.g. faults) (Rapolla et al. , 2002; Cella et al. , 2000; Tatchum et al. , 2011). The horizontal derivative method (Cordell and Grauch, 1985) represents one of the most used derivative approaches of potential fields, as it allows locating the horizontal position of the density or magnetization boundaries (Paul and Goodacre, 1984; McGrath, 1991; Rapolla et al. , 2002; Cella et al. , 2010; Stavred and Reid, 2010; Tatchum et al. , 2011). The expression for the horizontal derivative of a potential field M , in the space domain ( x,y ) is (2) The horizontal derivative has its maxima in correspondence of the lateral boundaries of the source anomaly. This supplementary information cannot be, of course, straightaway 144 GNGTS 2013 S essione 3.2