2. The Surveys

2.1 The use of geophysics in Scandinavian archaeology

While archaeological geophysics in some European countries, particularly the UK, developed rapidly in the 1980s and 1990s alongside the move towards developer-funded archaeology (Gaffney 2008), this trend has been relatively slow to take root in Scandinavia. A number of factors, both specific and general, have contributed to this situation (Gustavsen and Stamnes 2012; Viberg et al. 2011). In Denmark in particular, the institutional framework means that archaeological investigations are largely conducted through numerous small museums with limited resources to invest in specialised equipment or personnel. As a result, the application of different techniques has not always seemed justifiable and there has been a tendency toward conservatism, fuelled by concerns that such methods would divert already stretched resources from traditional excavations. Along with some pilot surveys that produced indifferent results, this has contributed to vicious circles of over-expectation and scepticism that hampered widespread application of these techniques in Scandinavia before the last decade or so.

Additionally, parts of Northern Europe suffer from combinations of geological and pedological circumstances that can make geophysical data difficult to interpret and archaeological features – often intangible – difficult to detect. For example, a programme of magnetometer and GPR survey over c. 2ha conducted in 2000 in advance of excavation at the early Viking trading site of Kaupang, Norway, identified a number of anomalies of anthropogenic origin, but extensive interpretation was not carried out owing to the difficulties of background geological noise and it was concluded that Norwegian geology was – at least in this case – unsuitable for geophysical investigation (Skre 2007).

Denmark has a comparatively well-established geological/hydrological geophysics sector, deploying marine, terrestrial and aerial techniques primarily with the aims of detecting minerals, hydrocarbons or groundwater; however, the application of terrestrial techniques on an archaeological scale is much less frequent. Geophysical prospection techniques were applied to archaeology in Denmark as early as 1956, with the use of a hand-held mine detector to scan for large Iron Age weapon sacrifices at Illerup Ådal (Andersen 1956). In 1965 the first archaeological magnetometer surveys conducted in Denmark were carried out on the Roman Age iron-smelting site of Drengsted in Southern Jutland (Abrahamsen 1965), but this technique has been applied infrequently since. Since the 1990s Tatyana Smekalova has conducted a large number of surveys on a range of site types across Denmark, as well as in other Scandinavian countries. The results of these surveys, conducted using Proton, Caesium, Overhauser and Fluxgate magnetometers, demonstrate that in general Danish geology and agricultural conditions are conducive to such survey, which provide a useful means of adding to our knowledge and understanding of archaeological sites (Smekalova et al 2008).

In recent years, developments in technology as well as interpretational expertise have facilitated the investigation of a number of important Iron- and Viking-age sites in Scandinavia and Northern Germany, using magnetometer and GPR survey to map the nature and extent of settlements, and identify individual features. In 2002 an area of 29ha over the Viking-age town of Hedeby, North Germany, was subjected to magnetometer survey, with spectacular results that demonstrated the complexity of the settlement within the ramparts and allowed identification of activity zonation (Hilberg 2007). The success of this technique on sites of this type is reinforced by more recent surveys from the nearby settlement of Füsing (Dobat 2010), and survey at the circular rampart of Tinnumburg, Schleswig Holstein, further supports the use of magnetometer results over fortress structures in this region (Segschneider 2009). Geophysical techniques are also being used to investigate and locate particular archaeological features that leave distinctive magnetic signatures: the effective use of magnetometry to locate SFBs (sunken-featured buildings) has been demonstrated on Iron- and Viking-age sites near Ribe, Jutland (Feveile et al. 2006), while evidence of high-temperature craft working has been detected on Gotland using similar methods (Gustafsson and Viberg 2012).

Increases in computing capability and instrument precision are now making landscape-scale surveys possible, driven largely by international partnerships and funding. The Ludwig Boltzmann Institute for Archaeological Prospection and Virtual Archaeology, for example, has forged partnerships between institutions in Sweden and Norway and long-established geophysics centres elsewhere in Europe, through which intensive investigations at key Iron- and Viking-age sites such as Birka (Trinks et al. 2007; 2010, Uppåkra (summarised in Viberg et al. 2011), and Gokstad (Jan Bill, pers. comm.) have taken place. Numerous anomalies from these sites have been interpreted as previously unknown features including buildings (at least one of which was burnt), cooking pits, hearths, a rampart, and paths (Voss and Smekalova 2007; Trinks et al. 2010). High-resolution GPR survey at Birka has facilitated the identification of individual postholes of 20cm diameter and probable graves, in addition to providing sufficient detail to offer a relative chronology of features (Trinks et al. 2007; Viberg et al. 2011). In Denmark the situation is also beginning to change with, for instance, the recent use of geophysical and geochemical methods to investigate the pre-Viking and Viking-age harbour site of Stavnsager (Loveluck and Salmon 2011), and a large-scale project by Moesgaard Museum to investigate Neolithic enclosures using GPR (Lutz Klassen, pers. comm.).

Within archaeological geophysics, recent years have seen an increasing tendency towards integrated survey, the obvious benefits of which revolve around the measurement of multiple physical properties of the ground in order to obtain complementary datasets. A number of teams have applied geophysical techniques at Uppåkra, and work has included fluxgate magnetometer, electrical resistivity, Vertical Electrical Sounding, GPR and electromagnetic surveys, in attempts to refine interpretations (Larsson 2001). Elsewhere, work at Sorte Muld, on Bornholm, illustrates the importance of multiple datasets: while geological conditions resulted in poor magnetic contrasts, a number of anomalies located in close proximity to a high number of gold finds and interpreted as a temple building could be more confidently identified as such from the corresponding radar data (Stümpel 2010). Investigations at Stavnsager, a pre-Viking and Viking-age harbour and settlement in east Jutland, are applying such approaches in an attempt to pursue a 'geo-archaeological dialogue' (Loveluck and Salmon 2011); in this case, targeted magnetometer, resistivity and GPR surveys have been undertaken alongside geochemical analyses and targeted excavation/ground truthing in order to identify occupation over a 100ha area. The integrated use of geophysics, both prior to and as an extension of excavation, has thus provided spatial and chronological insights.

Although the role of geophysics in archaeology has traditionally been one of mapping archaeology as a precursor to excavation (Conyers 2010), over the last decade, the rapid refinement of traditional data-capture techniques, the development of new technical analytical possibilities, and the increasing collective experience of archaeological geophysicists, are now making it possible to use such methods to approach archaeological and anthropological questions directly (Conyers and Leckebusch 2010; Kvamme 2003). It is increasingly being shown that geophysical survey lends itself in particular to investigations of the use and reuse of space, development/continuity of occupation, region-specific settlement characteristics such as building techniques, and, therefore, the social, economic and political characteristics of the inhabitants.