Truncation Filtering Methodology: Input gravity data and pattern matching

Abstract

The compilation of proper input gravity data for the Truncation Filtering Methodology (TFM) from observed gravity is discussed. The aim of the TFM interpretation is to determine the anomalous density distribution, or at least some of its characteristics, below the earth's surface in a studied region. It implies that the input data must be equal to the gravity effect (attraction) of all such anomalous masses of interest. Furthermore, the TFM requires that the input gravity data be given on a level reference surface, the position of which is further constrained by the requirement to stay outside all the anomalous masses, hence above all the terrain, in order to avoid downward continuation through anomalous masses. Such a requirement is imposed by the fact, that the TFM is a pattern recognition technique and the knowledge of patterns comes from synthetic modeling on a level surface without topography. Consequently the requirements imply that the input data needed are the gravity disturbances, corrected for the effects of topography and bathymetry, harmonically upward continued to a level surface tightly enveloping the topo-surface in the area of interest. Numerical procedures and several approximations in compiling such data are discussed.