Two-dimensional inverse filtering for the rectification of the magnetic gradiometry signal

Autor: Gregory N. Tsokas, Panagiotis Tsourlos, R. O. Hansen, Georgios Tassis, C. B. Papazachos
Rok vydání: 2007
Předmět:
Zdroj: Near Surface Geophysics. 6:113-122
ISSN: 1873-0604
1569-4445
DOI: 10.3997/1873-0604.2007037
Popis: The magnetic difference, the quantity measured by magnetic gradiometers, is considered to be the convolution between a function that controls the anomaly pattern and another that controls the strength signal. These are called shape and amplitude functions, respectively. They are distinct and analytically determined; thus, after the assessment of a suitable model, its shape function can be inverted to serve as a filter. If this filter is next convolved with the measured field, a series of amplitude functions is recovered, provided that the subsurface structures can be simulated by a combination of a number of models similar to the one whose anomaly was inverted. The recovered series is essentially the subsurface distribution of the amplitude function. Alternatively, the scheme can be viewed as a transformation of the original field of magnetic differences. The signals, transformed in this context, comprise rectified monopolar versions of the originals, positioned directly above the centre of the targets. Furthermore, their amplitude is a measure of the magnetization of the targets. Thus, the new anomalies can be viewed as a kind of magnetization or susceptibility mapping. Inversion filters are computed in the Wiener mode by inverting the shape function of the anomalies caused by vertical sided finite prisms. This particular model is appropriate for a wide variety of targets that are commonly met in archaeological prospecting. Converting the magnetic signal from the dual lobe pattern to single monopolar anomalies cause them to resemble the usual outcome of resistivity exploration. In this context, this work aims to form a scheme that makes the magnetic gradiometry outcome directly comparable to the outcome of resistivity mapping. Furthermore, it can be applied without contradicting or excluding any other processing operation. The method is implemented in a FORTRAN program that is reasonably user- friendly. The efficiency of the scheme is tested both on synthetic and real data sets.
Databáze: OpenAIRE
Externí odkaz: