A multi‐parameter approach for recognition of anthropogenic noise in aeromagnetic data collected over populated areas: Erzgebirge, Germany. (27th January 2022)

• Record Type: Journal Article
• Authors: Morris, William; Ugalde, Hernan; Kirsch, Moritz; Gloaguen, Richard; Ibs‐von Seht, Malte; Siemon, Bernhard; Meyer, Uwe
• Is Part Of: Geophysical prospecting. Volume 70:Number 3(2022)
• Page Start: 593
• Page End: 617
• Publication Date: 2022-01-27
• Subjects: Data processing -- Magnetics -- Noise; Prospecting -- Geophysical methods -- Periodicals; 622.15
• DOI: 10.1111/1365-2478.13179
• Languages: English
• ISSNs: 0016-8025

Abstract

ABSTRACT: Aeromagnetic data are routinely acquired by mineral exploration programmes. The objective is to obtain a raster image of the spatial variations of magnetic field intensity; these variations are associated with mineralogical variations in the subsurface. When the survey is conducted in a populated area, much of the signal, however, may be associated with anthropogenic sources such as buildings and roads. Identification and minimization of the anthropogenic‐related signal then are essential to derive a useful product for geological mapping. In this work, we examine a scalar magnetic dataset from Geyer, Saxony, and we apply five approaches for locating regions of anomalous anthropogenic signal: signal amplitude, absolute fourth difference, signal standard deviation, enhanced horizontal gradient and curvedness. All are shown to produce similar responses, and the summation of the five results compares favourably with the standard Keating kimberlite (circular anomaly) approach for detecting anthropogenic signals. Complications arise when geological features produce signals of similar amplitude to anthropogenic sources. Differentiating the probable origin of any specific pattern can be assessed by using a 2D shape index and increased flight height. Verification of an anthropogenic anomaly is achieved by comparison of anomalous solution grids with Geographic Information System‐based reference data.