Geophysical explorations of archaeological shell matrix sites: evaluating geophysical techniques in determining the boundaries, structure and volume of buried shell deposits

Kenady, Selene L. (2016) Geophysical explorations of archaeological shell matrix sites: evaluating geophysical techniques in determining the boundaries, structure and volume of buried shell deposits. PhD thesis, James Cook University.

PDF (Thesis)
Download (12MB) | Preview
View at Publisher Website:


Shell middens are a significant component of the Australian coastal archaeological record, however, they are notoriously difficult to research. Shell matrix sites are often large and structurally heterogeneous with complex formation histories. For large stratified shell matrix sites, the majority of the deposits are buried making the design of appropriate and representative sampling regimes challenging, as short of total excavation the population from which the sample was taken will never be fully understood. This study aims to address these sampling issues through the application of geophysical survey techniques to shell matrix sites, and by developing novel methods for creating three-dimensional models and volume estimates of buried shell deposits from these survey results.

An extensive literature review found only 22 published papers (representing 15 archaeological case studies) have applied geophysical surveys to shell matrix sites. Not one of these studies used geophysical methods to calculate volume estimates of the buried matrices or to create three-dimensional models of the deposits. Outside of archaeology, there have been studies (typically in glacial load research) which have attempted to create volume estimations from geophysical survey results. These studies have, however, typically not included ground truthing of results or calculated error values for the volume estimates.

This research has three primary aims: (1) to delineate and map buried shell matrix deposits in tropical Australian contexts; (2) to establish methods for transforming these survey results into volume estimates and three-dimensional models of the deposits; and (3) is to test the accuracy of these models and estimates. To achieve these aims two geophysical methods were chosen and compared; ground-penetrating radar (GPR) and electrical resistivity. These two geophysical methods were chosen, on the basis of the literature review, as being the most appropriate methods to meet the research aims. Both survey methods were employed under field and experimental conditions. Survey results were processed and then exported to Esri's ArcGIS suite of software for further processing, to create the three-dimensional models and volume estimates. Results from this modelling were then compared to the in-ground deposits to test their accuracy.

Survey and modelling results for the buried shell matrix deposits varied between geophysical method, and were dependent on the environmental conditions present on site. The electrical resistivity could not differentiate shell material from sand, but could differentiate shell from an organic-rich sediment. The GPR produced clearer, easier to interpret results under drier conditions, while the electrical resistivity produced them under wetter conditions. The modelled results showed more accurate three-dimensional representations of buried shell matrices could be created from the GPR, rather than the electrical resistivity surveys. Similarly, the volume calculations were highly accurate when based on GPR survey data, with an error margin on the estimates of 16%±11%, though it was found that small misinterpretations of the results can easily produce errors in excess of 50%. Volume calculations based on the electrical resistivity data were less accurate than the GPR and varied significantly depending on how the results were interpreted, meaning their overall error margin was significantly higher at 50%±29%. The geophysical survey results for this research also provided a greater understanding of the palaeolandscape on which the shell matrix at the field site was deposited.

In order to create accurate accounts of the archaeological record of coastal Australia it is vital that improved methods for characterising the variability of shell matrix sites are explored. The current research addressed this issue by evaluating the capabilities of two geophysical survey techniques in investigating buried shell matrices, and by developing methods for transforming the survey results into three-dimensional models and volume estimates. These methods provide a way to greatly improve sampling regimes in shell matrix research by providing an understanding of the buried deposits before excavation takes place. The methods also provide information in their own right, allowing for a better understanding of the size and shape of buried matrices, and the palaeolandscapes on which they were deposited.

Item ID: 50093
Item Type: Thesis (PhD)
Keywords: archaeology, Australian, electrical resistivity, geophysics, ground-penetrating radar, shell matrix
Date Deposited: 05 Sep 2017 04:51
FoR Codes: 21 HISTORY AND ARCHAEOLOGY > 2101 Archaeology > 210101 Aboriginal and Torres Strait Islander Archaeology @ 25%
21 HISTORY AND ARCHAEOLOGY > 2101 Archaeology > 210102 Archaeological Science @ 50%
04 EARTH SCIENCES > 0404 Geophysics > 040401 Electrical and Electromagnetic Methods in Geophysics @ 25%
SEO Codes: 95 CULTURAL UNDERSTANDING > 9505 Understanding Past Societies > 950503 Understanding Australias Past @ 40%
97 EXPANDING KNOWLEDGE > 970121 Expanding Knowledge in History and Archaeology @ 50%
97 EXPANDING KNOWLEDGE > 970104 Expanding Knowledge in the Earth Sciences @ 10%
Downloads: Total: 32
Last 12 Months: 12
More Statistics

Actions (Repository Staff Only)

Item Control Page Item Control Page