Advantages to Using the Pseudogravity Transformation to Aid Edge Detection of Total Field Archaeomagnetic Datasets

Abstract

Magnetic surveys are becoming increasingly common on archaeological sites due to the amount of data that can be collected rapidly in a non-invasive manner. Owing to their relatively low cost compared with excavation they commonly provide the only dataset that covers an entire archaeological site, which can then be used to target other surveys and excavations to areas of interest. Typically interpretation is done visually using optimized images of the raw data, which when dealing with large datasets can be time-consuming and subjective. Various derivative based methods have been developed recently to aid the interpretation of magnetic data. A particular use for these techniques is to locate the edges of subsurface magnetic bodies, and their use is gaining popularity in aeromagnetic regional and mineral exploration surveys. Despite this they are rarely used in archaeological survey interpretation. This probably is due to the particular challenges posed as a result of the low amplitude and high-wavenumber content of archaeomagnetic surveys, where features of interest are often only slightly above the noise level. Here, it is demonstrated that when derivative-based methods are applied directly to total-field data, the high-wavenumber components of the data are amplified, making datasets difficult to interpret and often proving less useful than the total-field dataset alone. The pseudogravity transformation is a readily available tool for suppressing this bias to the high-wavenumber features and providing derivative-based results with a power spectrum comparable to the original total-field response, but with all the qualities to enhance interpretation that are obtained from using the derivative methods.

Enhancing signals from buried Roman structures in Magnetometer data by combining continuous wavelet transform and tensor voting

“…Figure shows the results obtained by using the Euler deconvolution (Figure a), the ridge crests of the horizontal gradient (Figure c) (Cheyney et al, ) and theta map (Figure e) (Cheyney et al, ). The ridge crests of the edge‐filtered data are calculated by the method described in Hansen and Deridder ().…”

“…A disadvantage of derivative‐based filters is their sensitivity to noise. Cheyney, Hill, and Linford () propose to apply the pseudogravity transform on magnetic data before using edge filters. It suppresses high‐wavenumber components and, as a result, smooths the data.…”

Assessment of the Effect of Instrumental and Discretization Errors on Derivative‐based Potential‐field Geophysical Methods

“…In an attempt to reduce the high wavenumber content of the derivative‐based results, Cheyney et al . () proposed the use of pseudogravity transform before the computation of the derivatives.…”

“…The location of these edges may be more or less accurate depending on the particular application setting. However, the use of special derivative functions such as the so‐called ‘theta map’ (Wijns et al ., ) and ‘tilt angle’ (Miller and Singh, ; Verduzco et al ., ; Salem et al ., ) equalize the response of both shallow‐ and deep‐sited structures, acting like an automatic gain control (AGC) filter (Cheyney et al ., ).…”

Use of Edge Delineating Methods in Interpreting Magnetic Archaeological Prospection Data

“…Cheyney et al . (, ) have also presented similar work with the same objectives as this research. All the works on edge detection of buried structures can be considered as targeting towards the same direction (e.g.…”