Manual approaches to the removal of cultural noise from high-resolution aeromagnetic data acquired over hightly developed areas

Abstract

High-resolution airborne magnetic surveys in highly developed areas are prone to the problems of cultural noise or anthropogenic signal. This ‘noise’ may be defined as any part of the measured magnetic field that is derived from ‘non-geological’ sources that are not of interest to the study. In many surveys the ‘cultural’ component of the magnetic field can have significant amplitude and may mask features of geological interest, which can be of very low amplitude. The amount of cultural noise encountered tends to reflect the degree of human development in an area. Previous reports define cultural noise as singular, isolated anomalies (Muszala et al. 2001) or long linear magnetic features following pipelines (Gay 1986; Wilson 1997) or power-lines (Pearson 1996, Gharibi & Pedersen 2000). Several attempts have been made at automated removal of cultural noise, including successful application of wavelet analysis (Fedi et al. 2003). Cultural noise is highly troublesome in densely populated areas, such as large parts of the United Kingdom. Traditionally, magnetic surveys have been flown at increased altitudes in areas predicted to yield a high cultural noise content, or have not been flown at all. However, modern surveys are increasingly being conducted in urban and highly developed areas, as this is where water-resource, land-use and environmental problems are the greatest. This article aims to illustrate the level of cultural noise encountered during a high-resolution aeromagnetic survey of the English Midlands and the methods adopted to remove it. Frequency-based methods of filtering out cultural features were tried initially, but with limited success. The amplitude and wavelength characteristics of cultural noise and anomalies reflecting shallow geology were frequently similar. This spectral overlap presented a major obstacle to the numerical or automated removal of cultural noise and the failure of these existing methodologies resulted in our experimentation with manual approaches to deculturing. The two methods that were investigated both involve the identification of possible cultural sources of suspect anomalies on either in-flight video film or UK Ordnance Survey (OS) topography maps (at both 1:50 000 and 1:25 000 scales). Manual deculturing methods are traditionally viewed as slow, labour intensive and therefore costly. They can also introduce network levelling and gridding problems. The aims of our experimentation were to determine how effective the manual processes are at removing cultural noise and to produce a ‘clean’ baseline data set against which the products of quicker automatic and semi-automatic intelligent routines of deculturing can be compared. While little new insight is offered in this work, it is felt important to communicate the scale of the problem encountered and to show that by applying considerable care, it is possible to attain high-quality data.