Printer Friendly

Magnetometer surveys in archaeological research in Papua New Guinea: Keveoki 1, Gulf Province.


A magnetometer survey was conducted on the abandoned village site of Keveoki 1, near the Vailala River, Gulf Province, PNG. The survey, using a single sensor proton precession magnetometer, was successful in locating and defining the boundaries of areas confirmed by excavation to contain dense assemblages of pottery. The combination of geophysical and excavation results allowed a broader understanding of the spatial distribution of human occupation at Keveoki 1 than would have been possible based on excavation or visual field walking alone. We suggest this technique should be applied more regularly.

Keywords: magnetometry, archaeological geophysics, Papua New Guinea, pottery


Archaeological geophysical prospection techniques have not previously been applied as part of archaeological investigations in Papua New Guinea (PNG), despite an extensive history of archaeological research in this area (e.g. Bulmer 1978; Frankel and Vanderwal 1985; White and O'Connell 1982). In part, this deficiency may be explained by the perceived high cost of geophysical survey as well as the difficulties associated with operating and transporting electronic equipment to the often remote, extremely rugged, wet tropical and inaccessible archaeological sites of the region. Nevertheless geophysical techniques have a demonstrated history of making an important contribution to archaeological investigations world-wide (e.g. Witten 2006; Conyers 2004; Gaffney and Gater 2003) and have the potential to answer important archaeological questions in PNG also. In particular, they have the potential to extend site information beyond the limited spatial extent usually obtained through excavation, and thus promise to enable understandings of village sites as spatially extensive landscapes rather than more restricted spatial nodes (Kvamme 2003). This is particularly apt for PNG where thick vegetation and swampy conditions can make site discovery through more conventional field walking very difficult.

The archaeological record in many coastal parts of PNG is particularly amenable to geophysical investigations because here can be found extensive sites with dense ceramic deposits as well as numerous sub-surface structural features such as postholes, human burials and earth ovens. Since electromagnetic induction (EMI) and magnetic susceptibility in particular can directly detect pottery (Clark 1990) as well as the remnants of burning (Linford and Canti 2001) and anthropogenically-induced microbial activity (Linford 2004), geophysical prospecting evidently has great potential in such archaeological contexts. Other techniques, such as ground penetrating radar (GPR) (Conyers 2004) and direct current resistivity (Witten (2006) may find less regular application in this area, but could contribute where favourable site conditions exist.

Potential targets for archaeological prospection in PNG

While a variety of archaeological sites exist in PNG including rockshelter sites, coastal middens and agricultural landscapes, all of which have considerable potential for the application of geophysical techniques, here we focus our attention on the archaeological expression of an ancient village site from swampy southern PNG lowlands.

Recent villages of the Vailala River area--as of other areas of PNG also--typically housed a few dozen to a few thousand people, and contained from a handful to dozens of wooden and thatched houses. Some houses were particularly grand in size and reputation, such as the men's longhouses that could be over 100 m long and that housed sacred and secret ritual objects. Houses of all kinds were typically raised 1 m or more above the ground. It was common for family members to be buried in graves beneath individual houses.

Geophysical prospection techniques are well suited to investigating the spatiality of village organization in this part of PNG, and in doing so tracing archaeological expressions of local social institutions as recorded in the ethnographic literature. For example, the residents of Epemeavo and Kea Kea villages to the east of the Vailala River trace their ancestry to the upper waters of the Vailala and Purari rivers. Oral origin histories describe a series of 'halting places' occupied as people moved generally southwards toward the coast (e.g. David et al. 2009; Holmes 1903). This raises the question of whether ethnographically-described village layouts can be traced further inland and back in time. Delineating smaller domestic dwellings from small villages or temporary residences across the landscape would be difficult archaeologically, as these were more or less randomly scattered and regularly rebuilt. However, the orientation and utility of men's longhouses (eravo), especially common among the larger, coastal villages of this area, can be more clearly recognised. The eravo faces the sea, and is juxtaposed by two smaller baupo eravo facing inland. Outside times of ceremonial performance, these were places where men rested, socialised or entertained visitors (e.g. Williams 1940: 3-6). Eravo are not places where all the communal activities of day to day life took place, but rather in their positioning and hierarchical significance within village layouts signalled specific forms of social organisation involving gender and age relations; ceremonial procedures, obligations and schedules; and structures of ritual responsibility based around the clan and level of seniority which, in its archaeologically observable material expression (e.g. posthole distributions), can be expected to leave a unique material signature. However, such structures have not yet been recovered archaeologically.

Archaeological assemblages from village sites across this region are characterised by pottery, stone artefacts, bone, charcoal, abundant postholes, earth ovens, human burials as well as the geochemical signatures of occupation. These features have variable potential for detection using geophysical techniques, the most problematic being stone artefacts and skeletal materials. While large accumulations of stone artefacts might in principle be detectable using geophysical techniques such as GPR, EMI or direct current resistivity, their small individual size would make them very difficult to reliably detect in most cases. Further, it would be impossible to differentiate between worked and unworked stone with these methods.

More promising is the detection of skeletal material, particularly using ground penetrating radar (Conyers 2006, Ruffell and McKinley 2005). Most informed studies now consider that the direct detection of skeletal material under field conditions is impossible and that the best targets are the disturbance to the soil created through the act of burying the corpse or material culture items (such as coffins) associated with burials. For this method to be successful, it relies on minimal post-interment disturbance to the grave fill and surrounding stratigraphy (Conyers 2006: 67). In the tropical environments of PNG where jungle quickly overtakes any open ground, this degree of preservation is unlikely. An alternative approach is to detect the magnetic enhancement resulting from burial rituals involving fire or the use of ochre (Moffat et al. 2010; Wallis et al. 2008) which may have some potential for use in this region.

The target most amenable to geophysical investigation in much of PNG is increased magnetism caused by subsurface pottery, fire and other anthropogenic mechanisms of magnetic mineral formation. Of these pottery is, in general, the most likely to yield information about the chronology and material culture of a particular site. The increase in the magnetism caused by these features results from the cumulative contribution of both thermoremanence and induced magnetism, formed through different processes (Tauxe 2002) which (in this case) result from human activity.

Thermoremanence is acquired when iron oxide-rich materials (such as clay) are heated above the Curie point (578[degrees]C for magnetite and 578-675[degrees]C for maghemite; Schmidt 2007: 23), when the iron minerals are demagnetised then remagnetised en-masse in line with the earth's magnetic field when they cool (Clark 1990: 65).

The intensity of magnetic enhancement is greatest when the affected materials retain their spatial relationship, such as in an in situ kiln. When fired materials are found in different orientations relative to their thermally created magnetic field (such as in a collection of fired bricks in a wall) their directions of magnetisation are dispersed and the cumulative effect can be reduced (Bevan 1994). Despite this, the incorporation of broken ceramics and bricks into the soil can result in an increased magnetic signature (Weston 2002).

Induced magnetism is acquired by a complex series of processes first described by Le Borgne (1955), where iron minerals are reduced and then oxidised to form the more magnetically susceptible forms magnetite and maghemite. These processes, summarised by Aspinall et al. (2008: 22-26), include the heating of iron materials in reducing conditions; the microbial creation of reducing conditions and bacteria which internally reduce these minerals all rely on the abundant presence of organic materials.

The degree of the potential increase in the magnetism of the soil is controlled by the concentration of iron oxides present (Tite 1972), which can be measured experimentally (Crowther 2003). It is also possible to experimentally distinguish the enhancement in magnetic susceptibility caused by fire from that produced through weathering and soil formation processes (Oldfield and Crowther 2007).

Both of these mechanisms of magnetic enhancement have been the subject of significant experimental work to isolate the effects of parameters such as temperature, fire duration and the lithology of the study area on the intensity of the created thermoremanence and induced magnetism (Linford and Canti 2001).

The increase in magnetism can be detected using passive (magnetometer) and active (EMI or magnetic susceptibility) methods, with the most appropriate method depending on the nature of the targets and the site conditions (Clark 1990). Anthropogenic enhancement of the magnetism of a site through these methods therefore provides a robust basis for geophysical survey in this region.

Keveoki 1

The site of Keveoki 1 is located 9 km east of the Vailala River and 1.2 km north of the coastal village of Kea Kea (Fig. 1). The site was identified on the basis of pottery sherds by Bruno David and Nick Araho (David et al. 2008). The hinterland region behind the villages of Kea Kea and Epemeavo are interspersed with drainage ditches manually dug by local villagers to allow garden cultivation. These drains average approximately 1.5 m in depth and cut through old landscapes and shorelines, exposing archaeological material in the form of pottery sherds. Local villagers identify the pottery as being from old village sites, some remembered from the relatively recent historical past and others as part of oral traditions relating to ancestral villages. The Keveoki 1 pottery sherd concentration was thought by local clan members to be an old village but it was not known from oral traditions.

A single, small archaeological excavation was undertaken at Keveoki 1, where there was the highest density of pottery sherds in the bottom of the adjacent drainage ditches (Fig. 2). Pottery was visible along the length of the bottom of the drain for approximately 20 m, with minor densities along the top of the banks where mud had been excavated during ditch building, and in situ in the drain. Villagers from Epemeavo collected all but the smallest pieces from the ditch in the immediate vicinity of the excavation square using 2.1 mm mesh sieves, wet sieving inside the ditch and collecting all sherds above 20 mm in length. Densities were too great for a total sherd collection from the length of the ditch. The collected ceramics and those from the 50 x 50 cm excavation will be published elsewhere.

Geophysical survey methodology

Geophysical investigations were conducted using a Geometries G-856 proton precession magnetometer with data collected on a regular grid with one metre line and station spacing in areas of the Keveoki 1 site where the vegetation had been cleared over an area of approximately 500 [m.sup.2.] A data point was taken prior to the start of each survey line at a base station location situated outside the survey area to allow the calculation of a diurnal correction value, despite the absence of a second magnetometer.

Magnetometer investigations can be undertaken using a variety of instrumental configurations and positioning equipment. Common sensors for field based archaeological applications include proton precession, cesium vapour and fluxgate configured in single, gradiometer or multiple configurations, the relative utility of which are well summarised by Aspinall et al. (2008). Positioning information for the survey can be acquired by real time kinematic differential GPS, differential GPS, GPS, robotic electronic distance meter, electronic distance meter or survey tape depending on site conditions, required precision and accuracy of survey information, desired survey speed and available budget. The choice of a proton precession magnetometer and survey tape for this survey reflects the low cost of this configuration, the very low power requirements of these tools in an area where recharging was impossible during the survey, the small area allowing a comparatively slow methodology as well as the high quality of both positioning and geophysical data that can be obtained (e.g. Moffat and Raupp 2008).


Following acquisition, data were processed to remove erroneous points, diurnally corrected from a base station, gridded using Magpick software using a spline interpolation (Smith and Wessel 1990) and presented as a contour plot. This was overlain on a baseline/offset site plan (Burke and Smith 2004: 96) to assist in data interpretation (Fig. 3).

Geophysical survey results

The magnetometer data show a number of both discrete and diffuse anomalies which may correlate to anthropogenic features as originally reported in David et al. (2009: 12) and reanalysed here. The most distinct is a positive monopolar anomaly of up to approximately 17 nanoteslas (nT) above background centred on the area of the channel and excavation, in an area found to produce the most ceramic material (Fig. 3, feature M1). This anomaly continues, with a more diffuse boundary and slightly lower magnetic intensity of between 13-15 nT above background down river (Fig 3, feature M2). Also adjacent to feature M1 is a diffuse lobe, with moderately elevated magnetic intensity values in the range of 11-14 nT above background, which extends west for approximately 15 m with a maximum width of 8m (Fig. 3, feature M3). An additional lower magnetic intensity and geographically smaller positive monopolar anomaly with a maximum intensity of 13 nT above background is located approximately 12m to the east of feature M1 (Fig 3, feature M4). A negative monopolar anomaly with a minimum intensity of -10 nT below background is located to the south east of the principal anomaly (Fig 3, feature M5).



Discussion and conclusion

The results of the geophysical survey suggest that the channel has fortuitously been cut through the highest concentration of ceramic material in the area surveyed (Fig. 3, feature M1). This was the location of the excavation. The continuation of this anomaly downstream along this channel (Fig. 3, feature M2) is probably at least partly the result of fluvial transport of the pottery sherds, although part of this anomaly is located adjacent to the channel and so may be in situ. The eastward continuation of this feature (Fig. 3, feature M3) probably represents an additional, though less dense concentration of pottery material, as does the discrete smaller positive anomaly to the east (Fig. 3, feature M4). The anomaly to the southeast (Fig. 3, feature M5) probably does not represent a pottery accumulation. Its negative magnetic response requires direct investigation.

The results of the magnetometer survey suggest that despite the initial removal of the small number of pottery sherds analysed in David et al. (2009), a significant amount of material remains both in the creek bed and in situ in the creek bank. The continuation of feature M3 to the western edge of the survey area leaves open the possibility that further significant pottery deposits may be located outside the area surveyed.

The magnetometer data also suggest that significant regions of the survey area do not have extensive anthropogenic enhancement of their magnetic intensity and thus are unlikely to warrant further direct investigation. The implication is of a village site of limited spatial extent of some 25 m x 10 m, with outlying activity locations as would be expected in the case of individual houses lying on the edge of the village. We cannot comment on the possible extent of the village beyond the boundary of our survey area although note that feature M3 (Fig. 3) may continue to the west. This suggests that larger survey area may be useful in the future.

The Keveoki 1 results confirm the ability of magnetometer surveys to define the spatial extent of locations with rich buried ceramic assemblages in humid, tropical conditions. The significance of these findings is further enhanced by the fact that, historically, coastal villages in this general region are known to have tracked the rapidly migrating coastline. David et al. (2009:13) have reported that in this area the coastline has prograded on average 3 m per year since 500-550 years ago. The implication is that magnetometer surveys of a more reconnaissance nature can assist archaeologists to track settlement locations relative to shifting coastlines. Such investigations are particularly useful in exploring the dynamics of settlement systems during the late Holocene.


Thank you to the inhabitants of Kea Kea, Mai and Epemeavo villages, in particular Nao Hahi the owner of Keveoki 1, Rex Hahi and family for sheltering us during heavy rains, Moisen Iavi for his assistance in the field, and Sam Nao for telling us about the Keveoki 1 site and introducing us to Nao Hahi.

Our fieldwork wouldn't have been possible without the assistance of: Nao Hahi, Solomon Hahi, Oscar Hahi, Simeon Koivi, Auri Sam, George Sam, Gibson Kamug, Hareho Hoa, David Korea, Auks Hoahe, Huavape Erope, Jacob lope and Francis Kouri. Many thanks also to Francis Bafmatuk, Acting Deputy Director of the PNG National Museum and Art Gallery for support for this project, Timothy Akia of Radio Gulf for useful discussions and friendship, and the Deputy Provincial Administrator 'Mac' Orisuru Avai at Kerema. The staff of the PNG National Museum and Art Gallery provided support. Kara Rasmanis drafted Figure 1. This project was undertaken with the generous assistance of ARC Discovery grant and QEII Fellowship DP0877782 to BD and support from the USQ Public Memory Research Centre. Thank you to Peter White and Ross Stanger whose constructive editing contributed to the clarity of this paper.


Aspinall, A, Gaffney, C and Schmidt, A 2008. Magnetometry for Archaeologists. Lanham: AltaMira.

Bevan, BW 1994. The magnetic anomaly of a brick foundation. Archaeological Prospection 1 (2): 93-104.

Bulmer, SE 1978. Prehistoric culture change in the Port Moresby region. Unpublished PhD thesis, University of Papua New Guinea, Port Moresby.

Burke, H and Smith, C 2004. The Archaeologist's Field Handbook. Crows Nest: Allen and Unwin.

Clark, A 1990. Seeing Beneath the Soil: Prospecting Methods in Archaeology. London: BT Batsford Ltd.

Conyers, LB 2004. Ground-Penetrating Radar for Archaeology. Walnut Creek: Alta Mira Press.

Conyers, L.B 2006. Ground-penetrating radar techniques to discover and map historic graves. Historical Archaeology 40: 64-73.

Crowther, J 2003. Potential Magnetic Susceptibility and Fractional Conversion Studies of Archaeological Soils and Sediments. Archaeometry 45 (4): 685-701.

David, B, Araho, N, Kuaso, A, Moffat, I and Tapper, N 2008. The Upihoi Find: Wrecked Wooden Bevaia (Lagatoi) Hulls of Epemeavo Village, Gulf Province, Papua New Guinea. Australian Archaeology 66: 1-14.

David, B, Araho, A. Barker, B, Kuaso, A and Moffat, I 2009. Keveoki 1: Exploring the Hiri Ceramics Trade at a Short-Lived Village Site near the Vailala River, Papua New Guinea. Australian Archaeology 68:11-22.

Frankel, D and Vanderwal, R 1985. Prehistoric research in Papua New Guinea. Antiquity 59:113-116.

Gaffney, C and Gater, J 2003. Revealing the Buried Past: Geophysics for Archaeologists. Brimscombe Port: Tempus.

Groube, LM, Chappell, J, Muke, J, Price, D 1986. A 40,000 year-old human occupation site at Huon Peninsula, Papua New Guinea. Nature 324: 453-455.

Holmes, JH 1903. Notes on the Elema tribe of the Papuan Gulf. Journal of the Royal Anthropological Institute 33:125-134.

Kvamme, KL 2003. Geophysical surveys as landscape archaeology. American Antiquity, 68: 435-457.

Le Borgne, E 1955. Susceptibilite magnetique anomale du sol superficial. Annales de Geophysique 11 : 399-419.

Linford, NT 2004. Magnetic ghosts: mineral magnetic measurements on Roman and Anglo-Saxon graves. Archaeological Prospection 11 : 167-180.

Linford, NT and Canti, MG 2001. Geophysical Evidence for Fires in Antiquity: Preliminary Results from an Experimental Study: Paper Given at the EGS XX1V General Assembly in The Hague, April 1999. Archaeological Prospection 8:211-225.

Moffat, I, Wallis, LA, Hounslow, M, Niland, K, Domett, K and Trevorrow, G 2010, Geophysical prospection for Late Holocene burials in coastal environments: Possibilities and problems from a pilot study in South Australia. Geoarchaeology 25: 645-665.

Moffat, I and Raupp, J 2008. A New Approach to Investigating Shipwreck Sites in Littoral Environments: Multi-Technique Geophysical Investigations of Port Elliot, South Australia. Technical Briefs in Historical Archaeology, 3:1-7

Oldfield, F and Crowther, J 2007. Establishing fire incidence in temperate soils using magnetic measurements. Palaeogeography, Palaeoclimatology, Palaeoecology 249 (3-4): 362-369.

Ruffell, A and McKinley, J 2005. Forensic geoscience: applications of geology, geomorphology and geophysics to criminal investigations. Earth-Science Reviews 69: 235-247.

Schmidt, A 2007. Archaeology, Magnetic Methods, in Gubbins, D and Herrero-Bervera, E (eds) Encyclopedia of Geomagnetism and Paleomagnetism: 23-31. Dordrecht: Springer.

Smith, WHF and Wessel, P 1990. Gridding with continuous curvature splines in tension. Geophysics 55: 293-305.

Tauxe, L 2002. Paleomagnetic Principles and Practice. Dordrecht: Khiwer Academic Publishers.

Tite, MS 1972. The influence of geology of the magnetic susceptibility of soils on archaeological sites. Archaeometry 14: 229-236.

Wallis, LA, Moffat, I, Trevorrow, G and Massey, T, 2008. Archaeological geophysics in the Indigenous reburial process: A case study from Ngarrindjeri ruwe, South Australia, Antiquity 82: 750-760.

Weston, DG 2002. Soils and susceptibility: Aspects of thermally induced magnetism within the dynamic pedological system. Archaeological Prospection 9(4): 207-215.

White, JP and O'Connell, J 1982. A prehistory of Australia, New Guinea and Sahul. Sydney: Academic Press.

Williams, FE 1940. Drama of Orokolo: The social and ceremonial life of the Elema. Oxford University Press, London.

Witten, AJ 2006. Handbook of Geophysics and Archaeology. London: Equinox Publishing.

IM: Research School of Earth Sciences, The Australian National University, Canberra, ACT, 0200, Australia. and Department of Archaeology, Flinders University, GPO Box 2100, Adelaide, SA, 5001, Australia. BD, RS: Programme for Australian Indigenous Archaeology, School of Geography and Environmental Science, Monash University, Clayton, VIC, 3800, Australia. BB: School of Humanities and Communication, University of Southern Queensland, Toowoomba, QLD, 4350, Australia. AK: Papua New Guinea National Museum and Art Gallery, PO Box 5560, Boroko, NCD, Papua New Guinea. NA: itodao Research Systems Limited, P.O. Box 7773, Boroko, N.C.D., Papua New Guinea.

IM: Research School of Earth Sciences, The Australian National University, Canberra, ACT, 0200, Australia. and Department of Archaeology, Flinders University, GPO Box 2100, Adelaide, SA, 5001, Australia. BD, RS: Programme for Australian Indigenous Archaeology, School of Geography and Environmental Science, Monash University, Clayton, VIC, 3800, Australia. BB: School of Humanities and Communication, University of Southern Queensland, Toowoomba, QLD, 4350, Australia. AK: Papua New Guinea National Museum and Art Gallery, PO Box 5560, Boroko, NCD, Papua New Guinea. NA: itodao Research Systems Limited, P.O. Box 7773, Boroko, N.C.D., Papua New Guinea.
COPYRIGHT 2011 Blackwell Publishing Limited, a company of John Wiley & Sons, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Moffat, Ian; David, Bruno; Barker, Bryce; Kuaso, Alois; Skelly, Robert; Araho, Nick
Publication:Archaeology in Oceania
Article Type:Report
Geographic Code:8PAPU
Date:Apr 1, 2011
Previous Article:Temporal trends in prehistoric fishing in Palau, Micronesia over the last 1500 years.
Next Article:Mid Holocene maritime economy in the western Torres Strait.

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |