Geophysical and archaeological investigations of Baker’s Flat, a nineteenth century historic Irish site in South Australia

Kelsey M. Lowe, Susan Arthure, Lynley A. Wallis, Josh Feinberg

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


The Irish settlement of Baker’s Flat, located in the rural heart of South Australia near the town of Kapunda, was occupied from the mid-nineteenth century for about 90 years. Although little archaeological work has been carried out in Australia specifically on Irish communities, Baker’s Flat is of particular interest because it potentially operated as a traditional Irish clachan, an informal clustering of farm dwellings and outbuildings, and home in this instance to the Irish immigrants who worked in the nearby copper mine. The site was cleared for farming purposes in the 1950s, and little recordation of the dwellings and settlement exist today, aside from a single 1890s map. Owing to the demolition and landscape modification, it was unclear whether any intact subsurface deposits still existed. Therefore, this site was ideal for deploying two geophysical methods, ground-penetrating radar (GPR) and magnetic gradiometry, to assess the presence of subsurface remains and explore the spatial layout of the site. Our results, when compared with those obtained from surface surveys and targeted archaeological excavation, revealed numerous subsurface features and helped to confirm that Baker’s Flat was built in the style of a traditional Irish clachan. This study also demonstrated that magnetic gradiometer was the better geophysical method for mapping this settlement as the nature of the geology (shallow bedrock) and construction of the houses (built within the bedrock) limited the utility of GPR.

Original languageEnglish (US)
Article number33
JournalArchaeological and Anthropological Sciences
Issue number1
StatePublished - Jan 1 2020

Bibliographical note

Funding Information:
The authors thank Jordan Ralph and Simon O’Reilley for their assistance with the geophysical surveys—their help allowed us to complete the survey within the allocated timeframe. Dale Hampel generously provided access to the land. Rob Koch provided surveying support. Our thanks also to the Kapunda community for ongoing support, and especially Peter and Jenny Swann, Fr Mark Sexton and Andrew Philpott. A Flinders University staff and student field crew participated in the 2016 and 2017 excavations, and were integral to successful completion of the fieldwork, with special thanks to Heather Burke, Cherrie De Leiuen, Bob Stone and Meg Haas. Susan Arthure’s research is supported through an Australian Government Research Training Program Scholarship. 1 The Kapunda Copper Mine was Australia’s first successful metal mine and in recognition of its historical significance was inscribed on the SA Heritage Register in 1987. 2 These relative differences have significant implications for magnetic gradiometry at archaeological sites. The Q value, or Koenigsberg ratio, is the ratio of a rock’s remanent magnetisation (Mr), meaning the magnetic moment of substances when no magnetic field is applied to its induced magnetisation meaning the magnetic moment when a magnetic field is applied to a substance. In this study, we used the NRM intensity for measuring Mr. Since low Q value (< 0.5) samples are dominated by the induced magnetisation, their magnetic anomalies will appear as magnetic highs (positive) compared with more non-magnetic crusts. The amplitude of an induced anomaly increases with the magnetic mineral content. When Q values are > 2, the orientation of a sample’s NRM becomes essential. When Q equals 5, the dominant remanence can lead to large negative anomalies, especially when the NRM vector is at an obtuse angle to the Earth’s magnetic field direction. Assuming an ambient magnetic field intensity at Baker’s Flat is 58 μT (46 A/m), the Koenigsberger ratio for the two samples is significantly greater than 2 (see Table 1 ). As such, their strong magnetic remanence will likely contribute towards elevated anomalies with respect to surrounding soil and sediment with lower Koenigsberger ratios. The average sediments Q values are between 0.02 and 10 (Hunt et al. 1995 ). Since Baker’s Flat is less than 1000 years old, the remanence direction of the slag should be roughly parallel to the modern field, thereby generating large positive anomalies such as we see in the gradiometer map (see Fig.  6a ). The susceptibility as a function of temperature, which is good for detecting the magnetic minerals of the slag, was remarkably reversible on warming and cooling (Fig.  13a ). The Curie temperature (Tc) was slightly higher during warming (471 °C) than during the cooling phase (463 °C). The slight decrease on cooling may be associated with partial oxidation of the assemblage or with the relaxation of confining pressure on the mineral grains embedded in the rapidly quenched silicate glass. These temperatures are lower than the Curie temperature of pure magnetite (580 °C) (Evans and Heller 2003 ; Thompson and Oldfield 1986 ) and are more consistent with magnetite partially substituted with metals such as Titanium (Ti), Chromium (Cr), or Magnesium (Mg). The sample shows mostly a single domain-like behaviour with minor contributions from the pseudo-single domain (PSD) and multidomain (MD) magnetic grains (Dunlop 2002 ) (Fig.  13b ). Magnetic domains are small regions in which the magnetisation is uniform within a sample, but the magnetisation vector with each region differs from that of its neighbours (Evans and Heller 2003 ). Plots of the back-field derivative of the data show two overlapping coercivity distributions with peaks near 30 and 75 mT (Fig.  13c ).

Publisher Copyright:
© 2020, Springer-Verlag GmbH Germany, part of Springer Nature.


  • Australia historical archaeology
  • Baker’s Flat
  • GPR
  • Historic Irish site
  • Magnetic gradiometry


Dive into the research topics of 'Geophysical and archaeological investigations of Baker’s Flat, a nineteenth century historic Irish site in South Australia'. Together they form a unique fingerprint.

Cite this