Linear filters for solving the direct problem of potential fields

Abstract

One of the major advances of computer science in recent years is the introduction of parallel processors. The efficiency of such devices depends strongly on the symmetry of the algorithms implemented; probably the most efficient are fast Fourier transform‐type (FFT‐type) algorithms. It is possible to show that for these algorithms a SIMD (Single Instruction Multiple Data) processor reduces the number of operations from [Formula: see text] to [Formula: see text] (where N is the number of input data and P is the number of processors) provided that both N and P are integer powers of 2. It thus seems advantageous to reconsider the possible uses of the fast algorithms in geophysics, including those not successful on nonparallel computers (Bezvoda et al., 1986).

Quantitative interpretation of magnetic data over settlement structures by inverse modelling

“…This is equivalent to numerical integration of the formula for the magnetic anomaly due to a vertical prism using an open one-point Newton±Cotes formula for both horizontal integrations, and an extended trapezoidal rule for the integration along the vertical axis (Bezvoda et al, 1992). In this case, different weights should be assigned to the dipoles at both ends of every column, as follows from the extended trapezoidal rule (see Press et al, 1990, p. 116).…”