1 Introduction

Geophysical methods are an important component of geoarchaeological investigations due to their ability to non-invasively image the subsurface of an archaeological landscape . New developments in multi-sensor and positioning technology have facilitated the use of these methods over large areas, allowing archaeological questions to be addressed on a landscape scale. They are particularly useful in geoarchaeological investigations for defining site stratigraphy, mapping site disturbance and reconstructing palaeolandscapes.

Geophysical methods also make a significant contribution to archaeological investigations beyond geoarchaeology. Excellent detailed introductions to the application of specific geophysical techniques to archaeology in general are available for ground-penetrating radar (Manataki et al. 2015; Conyers 2013, 2015b), magnetometry (Armstrong and Kalayci 2015; Aspinall et al. 2008), resistivity (Schmidt 2013) and electromagnetic techniques (Simon and Moffat 2015).

2 An Overview of Geophysical Methods

2.1 Ground-Penetrating Radar

Ground-penetrating radar (GPR) is a widely used technique in geoarchaeology, which allows the detection of features in the subsurface that cause a change in the propagation of electromagnetic energy. A typical GPR includes a transmitting antenna (Tx) and a receiver antenna (Rx). The Tx transmits electromagnetic waves with a fixed frequency into the subsurface. The Rx receives the portion of the energy, which is reflected by variations in material properties of the subsurface and registers the amplitude of this response for mapping purposes (Fig. 14.1).

Fig. 14.1
figure 1

Signal paths between the transmitter (Tx) and the receiver (Rx) for two layers with different electrical properties (ε stands for the dielectric permittivity). The dashed yellow line indicates that part of the signal’s energy was refracted and then reflected (Annan 2009)

Electromagnetic wave propagation is complex as it includes reflections, refractions and diffractions caused by changes in relative dielectric permittivity (ε) and/or conductivity (σ). Variations in these values reflect geological and/or geomorphologic features (e.g. changes in lithology, faults, strike and dip of beds, cavities, bedding features, loose sediments vs. bedrock etc.), changes in hydrological conditions (degree of fluid saturation and water chemistry) or the presence of archaeological materials. GPR can be deployed with a variety of antenna frequencies allowing features on a wide range of scales to be imaged. In general, antennas with a frequency range of approximately 200–500 MHz are appropriate for most archaeological studies as they provide a depth of investigation up to about 3 m and an acceptable resolution (Fig. 14.2a) (Jol 1995; Conyers 2015a, b, pp. 14–15). Lower-frequency antennas can be used for deeper investigations (such as the location of geomorphic features), and higher frequency antennas are suitable for shallower, higher resolution surveys such as investigation of mortar thickness. Recent developments in GPR technology include multichannel systems (Fig. 14.2b) coupled with GPS navigation systems for fast surveys, ultra-wide-band antennas (Trinks et al. 2010) and stepped-frequency antennas (Linford et al. 2010). Drone-mounted GPR systems are also expected to become available within the coming years (Merz et al. 2015).

Fig. 14.2
figure 2

(a) Noggin Plus (Sensors & Software Ltd.) 250 MHz GPR system in a survey at the ancient theatre of Chersonessos in Crete. (b) MALA MIRA multi-antenna GPR configuration with eight 400 MHz antenna applied in a survey at Feres (Velestino), Central Greece

Usually, GPR surveys are undertaken on a grid of regularly spaced parallel lines, which are commonly post-processed together into a 3D volume. Radargrams represent the amplitude of the reflected signals as a function of the time taken for the electromagnetic wave to travel from the Tx to the Rx via the subsurface. When the velocity of propagation of the electromagnetic radiation within the ground can be estimated, it is possible to convert the time of arrival to a depth estimate. GPR data can be viewed as 1D soundings (traces), 2D profiles (radargrams), amplitude ‘slices’ through the 3D volume, 3D cubes or 3D representation of the signal amplitude. In general, amplitude slicing is more effective for mapping archaeological features with obvious linear geometry and a high amplitude response than for stratigraphic features which are often more geometrically complex. For the investigation of geoarchaeological research questions, the combined analysis of individual traces, 2D profiles and slice maps is most likely to be more effective than the exclusive use of slice maps (Conyers 2015a, b).

GPR applications in geoarchaeology range from determination of the depth to bedrock, mapping of palaeolandscape features (such as channels and shorelines, Fig. 14.3), locating of anthropogenic earthwork features like canals or moats and identification of soil disturbance due to habitation (Fig. 14.4). In addition, GPR allows mapping unmarked graves, which often can be identified through soil anomalies caused by excavation of tombs. Unfortunately, the high attenuation of radar energy in saltwater precludes the application of GPR in marine surveys, but it is suitable for use in limnic environments (Fig. 14.3).

Fig. 14.3
figure 3

(a) Mala X3M 100 Mhz GPR system being used for mapping sedimentary environments within a lake. Lake Alexandrina, South Australia. (b) A palaeochannel (see the yellow line) imaged in a 2D profile using a Mala X3M 250 MHz ground-penetrating radar from the Umbum Creek, Central Australia

Fig. 14.4
figure 4

(a) Results of GPR slices at about 60–70 cm below the surface resulting from Noggin Plus and MALA. (b) GPR systems from Demetrias archaeological site, Central Greece. Measurements were taken along transects 0.5 m apart

A major advantage of GPR for geoarchaeological studies is that it can image sedimentary structures, such as cross bedding. Thus, it is an effective technique to be applied in areas characterized by changes in depositional facies that are not accompanied by lithological changes. Moreover, as the GPR antenna can be moved continuously during data acquisition, the method allows faster data acquisition and higher survey speed compared to methods such as resistivity tomography (see below) that require stakes to be inserted in the ground. GPR also benefits from the shielding commonly placed around antennas, which reduces the interference from features on the surface (particularly metal) as opposed to other methods, such as electromagnetic induction or magnetometry, making it suitable for use in urban settings (Sarris and Papadopoulos 2012; Papadopoulos et al. 2009).

2.2 Electromagnetic Induction Methods

Electromagnetic induction (EMI) can be used to calculate the apparent conductivity and magnetic susceptibility of the subsurface. In the most common type of EM instruments, a transmitter coil generates a primary electromagnetic field for a given frequency (~8–90 kHz) (Fig. 14.5). The interaction of the field with subsurface elements generates an electrical current. This electrical current, in return, generates a secondary electromagnetic field which is sensed by a receiver coil in the instrument alongside the primary field. The magnitude of secondary field is divided into two orthogonal components: the quadrature and the in-phase. The quadrature provides information on apparent conductivity (in mS/m), and the in-phase is related to the magnetic susceptibility (emu) of materials. The depth of investigation is mainly controlled by the separation between the transmitter and the receiver coils (Fig. 14.6; Simon et al. 2015a, b).

Fig. 14.5
figure 5

Diagrammatic representation of the principle of operation of the electromagnetic induction methods

EM requires no direct contact with the ground surface allowing large areas to be surveyed relatively quickly (Fig. 14.6 and 14.7). Conductivity can be a good measure of changes in rock and soil units, because it is influenced by composition, the porosity, degree of water saturation and fluid chemistry. Magnetic susceptibility may provide information about erosion, pedogenesis and anthropogenic burning as all of these processes may form or concentrate magnetic minerals (French 2003).

Fig. 14.6
figure 6

GEM2 Geophex (a) and CMD Mini Explorer–CF Instruments (b) configurations and coil arrays for the investigation of the soil conductivity and soil magnetic susceptibility with respect to the coil separation and orientation of coils (horizontal HCP and vertical VCP)

Another subfield of the use of electromagnetic induction in geoarchaeology is the specific investigation of magnetic susceptibility by dedicated instruments. Corresponding instruments are more sensitive to changes in magnetic susceptibility than general-purpose electromagnetic induction instruments and can be used to survey on the surface, on archaeological sites and excavations and even down in drill holes (Dalan and Banerjee 1998). However, such applications are often complemented by more detailed magnetic measurements (Dalan 2007), as explained in the section on magnetic methods below.

Fig. 14.7
figure 7

Indicative results from the EMI survey (GEM2) at the Neolithic site of Almyros 2 at Thessaly, Greece representing the soil conductivity (a), which suggests a higher conductivity to the southern side of the tell (indicative of the flooding zone) and the magnetic susceptibility (b) that outline the surrounding ditches and some of the details of the structural remains of the core of the settlement (Sarris et al. 2015a)

2.3 Electrical Resistance Techniques

Electrical resistivity investigations work by measuring either the self-potential (passive) or the direct current resistivity or induced polarization (active) of the subsurface. The resistivity of geological and archaeological materials is controlled by their physical properties as well as their water chemistry and degree of saturation (Hecht and Fassbinder 2006).

Electrical resistivity surveys are undertaken between probes of a known separation. The larger the separation between the electrodes, the greater the penetration depth and the smaller the resolution. Various electrode configurations can be used, including the Twin, Wenner, Square, pole-dipole, dipole-dipole and Schlumberger arrays (Loke 2000; Fig. 14.8). Each configuration provides particular advantages for specific situations, depending on the size, geometry, orientation, depth and relative resistivity contrast between the target and background (Schmidt 2013). Multi-probe, wheeled and tractor-driven systems have also been developed for smooth and conductive landscapes.

Fig. 14.8
figure 8

Indicative electrode configurations used in a soil resistance survey. Blue arrows represent potential electrodes, and red arrows represent current electrodes. The distance between them varies according to the electrode array. In the mapping mode, either all or some of the electrodes move within the area of interest

Resistivity measurements can be carried out in various ways. Vertical electric soundings, in which the electrode separation is gradually increased at a fixed point, are used to obtain a model of the depth to bedrock at a fixed location. Resistivity mapping, the most common configuration for archaeological prospection, makes use of a fixed electrode separation which is moved as an array along parallel profiles within a grid. This approach produces a plan-view image of the distribution of the resistivity at a specific depth. Finally, multi-probe arrays [commonly referred to as electrical resistivity tomography (ERT) ] use multiple equidistant electrodes that measure both the lateral and vertical variations of the subsurface resistivity as a 2D profile (Fig. 14.9). Having a number of parallel ERT transects, it is possible to generate 3D volumetric images of the resistivity. Resistivity, especially when used in the ERT configuration, together with seismic techniques, is excellent for imaging changes in lithology and geology (Laigre et al. 2012; Scapozza and Laigre 2014). ERT is also valuable when either deep depositional targets (e.g. ancient ports covered by alluvial deposits, ditches and palaeochannels) or offshore archaeological features are the target of investigations (Sarris et al. 2014; Tonkov 2014; Simyrdanis et al. 2015).

Fig. 14.9
figure 9figure 9

(a) Geoscan Research Resistance meter (RM15) with multiplexer employing six electrodes (current electrodes As and potential electrodes Ms) at various configurations. (b) ERT extension of electrodes being used to map the area of the old Franciscan monastery of Rethymno (Mikrasiaton plaza) in Crete. (c) Stratigraphic results of an ERT transect measured above a fault (see the black line) within the suggested ancient port of Priniatikos Pyrgos at Istron, Crete (Sarris et al. 2014). (d) 2D representation of the 3D resistivity model shown as horizontal slices with increasing depth in the area of Mikrasiaton Square at the centre of the old town of Rethymno in Crete. The architectural relics of a Franciscan monastery become obvious towards the southern part of the surveyed area (Papadopoulos et al. 2008)

2.4 Magnetic Methods

Magnetic prospection (or magnetometry) works by measuring disturbances to the earth’s magnetic field caused by the presence of iron minerals. Magnetization of bodies consists of two components, namely, the induced and the remanent magnetization. While the former is created by the modern magnetic field, the latter is a result from previous magnetic fields. Magnetic susceptibility is an additional important parameter, which measures the degree to which a material becomes magnetized when an external magnetic field is applied. Magnetic prospection depends on several parameters, such as the degree and orientation of induced and remanent magnetization of bodies, their magnetic susceptibility, their volume (or mass) and their distance from the sensor.

Magnetic measurements can be collected with single or multiple sensors configured to measure usually either the total magnetic field intensity or the total or vertical magnetic gradient (gradiometer mode). Recent equipment developments have introduced multiple sensor array carts to archaeology which, when combined with high-quality positioning and navigation systems, facilitate extremely high-sample density measurements over large areas extremely quickly (Fig. 14.10).

Fig. 14.10
figure 10figure 10figure 10figure 10

(a) A single fluxgate gradiometer sensor of Bartington G601. (b) An eight-sensor fluxgate gradiometer array of SENSYS surveying the Neolithic tell of Perdika 2 at Central Greece. (c) Indicative results from the single sensor magnetic survey of the Agora of Sikyon in Peloponnese (Sarris et al. 2008). (d) Multisensory survey of Almyriotiki Neolithic tell at Central Greece. Results have been superimposed on a WorldView-2 satellite image (2 June 2012) (Sarris et al. 2015a)

Contrary to its widespread application for archaeological prospection, field magnetometry plays a rather little role within geoarchaeological investigations. This is mainly because it is rather suited for the direct detection of ferrous or burnt materials than for understanding the stratigraphic matrix that contains archaeological material. One promising application, though, is mapping the extent of pedogenic soil minerals, which is more effective using single sensor equipment instead of gradiometry equipment.

In contrast, laboratory magnetic measurements are very useful for understanding the depositional conditions at archaeological sites (Dalan and Banerjee 1998). Magnetic susceptibility, particularly the frequency dependant component, is increased by archaeological occupation both through heating and the addition of organic material to the soil. The surface distribution of magnetic susceptibility values provides a proxy for the spatial distribution of occupation, and, in combination with other magnetic properties [e.g. anhysteretic remanent magnetization (ARM) and saturation isothermal remanent magnetization (SIRM)], it is possible to infer the intensity of habitation or to determine the provenance of archaeological materials (Thompson and Oldfield 1986; Dearing 1994).

2.5 Acoustic Procedures

Acoustic measurements are undertaken by creating repeatable acoustical waves that enter the ground and are refracted or reflected by changes in density before returning to a series of geophones (Fig. 14.11). The study of the arrival times, amplitude and form of the returned wavefronts provides information about the stratigraphy of the subsurface and the presence of archaeological material.

Fig. 14.11
figure 11

Typical setup in a seismic survey. The seismic source is needed to generate controlled acoustical waves that are either reflected or refracted at the different media interfaces to be recorded by the geophones. Geophones convert the velocity of the registered waves into voltage

The most common acoustic techniques in archaeological research are seismic reflection and bathymetry, particularly for the reconstruction of underwater palaeolandscapes (i.e. Tizzard et al. 2015). The acoustic response of subsurface sediments measured by seismic reflection provides two- and three-dimensional representations of the geometry of sedimentary depositional units, their internal structures and their lithology. Seismic methods are particularly suited to aquatic surveys, as the acoustic properties of water allow the geophones to be used without direct coupling to the seafloor which greatly enhances survey speed compared to terrestrial surveys.

On land, seismic methods are useful for geoarchaeological investigations as their relatively deep depth of penetration is well suited to mapping large monumental structures and mounds or reconstructing the palaeotopography. Seismic refraction is particularly useful for determining the depth to bedrock within archaeological sites (Fig. 14.12). In contrast, seismic reflection is very valuable in terrestrial archaeology to map site stratigraphy, with surface wave techniques in particular showing great potential for wider use.

Fig. 14.12
figure 12

Results of a seismic refraction survey at the area of Priniatikos Pyrgos (Istron, E. Crete, Greece), where an assumed ancient port has been hypothesized. The area has been completely covered by alluvial deposits and conglomerates as a result of past landslides and tectonic activity. The image represents the depth to the bedrock which reaches levels of about 40 m below surface, superimposed on a panchromatic Ikonos image (30 July 2001). The arrows represent the seismic transects along which the geophones were placed. The seismic results have also been verified through ERT measurements (Sarris et al. 2014)

2.6 Other Geophysical and Geochemical Techniques

Microgravity measurements are carried out using sensitive devices that measure small variations in the Earth’s gravitational field. These alterations are usually caused by air-filled voids in the subsurface (e.g. tunnels, cavities, caves and architectural features; see Fig. 14.13). They show a significant contrast between mass density and the surrounding loose sediments or the parent rock.

Fig. 14.13
figure 13

(a) GPR radargram (taken with Noggin Plus GPR with 250 MHz antenna) above a Roman chamber tomb at the Kenchreai cemetery located at Koutsonglia Ridge, north of the harbour of Kenchreai in NW Peloponnese. The tomb has a diameter of 3.55 m and an interior height of about 2.5 m, located at a depth of 1.1 from the ground level. The amplitude of the GPR data is represented by red (positive reflection), blue (negative reflection) and green (minor dielectric constant differences between the media creating almost no reflective interfaces) colours. (b) Gravity residual anomalies recorded at the same transect using the Lacoste and Romberg model D land microgravitometer. The centre of the tomb chamber is located approximately at the middle of the transect (Sarris et al. 2007)

Gamma spectroscopy is another geophysical approach that measures the gamma rays emitted by radioactive materials. It has been widely used by soil scientists to map soil types through plotting ternary diagrams of K-Th-U concentration. Mapping soils with gamma spectroscopy is very effective, because the radiometric footprint of a soil reflects both the mineralogy of the parent rock and the effect of weathering. Even though this technique has not been widely used in archaeology, it qualifies well for validating soil mapping results obtained by complementary methods like electromagnetic induction.

Another approach is the mapping of novel and short-lived isotopes such as 137Cs and 241Am, which were deposited by the fallout from nuclear testing and accidents. The method provides an alternative for mapping disturbance on the basis that these isotopes are deposited on the surface and chemically immobile. Given that isotopes are distributed throughout the whole soil profile rather than just on the surface, a mixing of the soil layers can be presumed. Both Cs and Am approaches have the potential to assist geoarchaeological investigation by providing information about the sedimentary matrix of the archaeological site and determining if it has recently been disturbed by anthropogenic or natural processes.

Quite similar to isotope mapping is the chemical analysis of soils, which may indicate anthropogenic activity including agricultural and workshop activities, animal husbandry, construction of ditches, etc. Phosphate analysis, trace element analysis (Cu, Mn, Mo, Ca, Se, Sr, Zn, Cd, Cr, Cu, Ni, Pb) and assessment of chemical stability of organic chemical compounds are the most widely used approaches contributing to the recognition of former land-use practices and past human occupation in general. Calorimetric measurements, atomic absorption spectrometry, inductively coupled plasma mass spectrometry and gas chromatography-mass spectrometry are usually employed (Szostek et al. 2005; Price and Burton 2012; Manhita et al. 2014; Lauer et al. 2014)

3 Examples for Principal Applications

Geophysical methods have a wide range of application within archaeology in general, but their explicit use for solving geoarchaeological questions is principally related to three major aspects: the determination of sediment stratigraphy, the mapping of soil disturbance related to human occupation and the reconstruction of palaeolandscapes .

Sediment Stratigraphy

Understanding the stratigraphy of an archaeological site is crucial for effective excavation planning, palaeoenvironmental reconstruction and putting excavation results into context, particularly with regard to geochronology. Yet geophysical methods are mostly currently used to simply locate and map archaeological sites and not to determine key stratigraphic features. The most important information that can be obtained using geophysics includes depth to bedrock, thickness and lateral extent of stratigraphic features and the location of archaeological material with respect to the stratigraphy.

Depth to bedrock and geomorphology of the surface provide important information about past depositional conditions and are key for correctly locating archaeological excavations so as to resolve the most complete record of occupation. Besides GPR and ERT, which are the most suitable techniques for investigating the vertical stratigraphy of archaeological sites, on-site or down-hole magnetic susceptibility measurements also provide promising results. When it comes to horizontal mapping, EMI, GPR, soil resistivity, ERT and gamma spectroscopy are particularly suitable.

Stratigraphic Disturbance of Sediments and Soils

Soil stratigraphy (i.e. the development of recognizable soil horizons) typically takes significant time to develop, and, as a result, its disturbance can provide crucial information about the historical record of an archaeological site. This perturbation can be mapped geophysically through a number of techniques. Among others, alteration of soil properties due to mixing of different components can be investigated by GPR, EMI, resistivity or gamma spectrometry, while changes of soil porosity are shown best by ERT and GPR. The latter approach is particularly appropriate for the detection of unmarked graves, as in the absence or following the decay of a coffin; soil perturbation caused by the dig is the most evident feature to be detected by geophysical methods (Moffat 2015; Conyers 2006).

Reconstruction of Palaeolandscapes

Buried palaeolandscapes provide valuable information for the identification and the understanding of archaeological sites, particularly when surface features are absent. The reconstruction of palaeoenvironments allows an identification of areas with sediment accumulation (and, thus, prospective for preserving archaeological material) and areas characterized by landforms (such as rivers or shorelines) conducive to ancient land use. In addition, it can also provide information about past climatic conditions, which had an impact on human occupation history. Besides terrestrial palaeolandscape reconstructions , which are quite often based on multi-method approaches (e.g. by combining ERT, seismics, geochemical studies), aquatic palaeoenvironments that have been submerged due to sea level rise are primarily investigated by acoustic methods (Tizzard et al. 2015).

4 Final Remarks

Geophysical techniques can make a significant contribution to geoarchaeological investigations. They allow non-invasive and rapid imaging of archaeological landscapes and help answer scientific questions by considering a site integrally along with its environmental surroundings. The methodological spectrum includes the investigation of the site stratigraphy, the mapping of remnants of past human occupation and the reconstruction of palaeolandscapes . Each method has its own merits and limitations, and it is through the combination of the various techniques that various aspects of the archaeolandscapes can be addressed (Sarris, 2013, 2015) (Table 14.1).

Table 14.1 Strengths and weaknesses of geophysical methods in geoarchaeological surveys

It has to be noted though that geophysical prospection techniques cannot and shall not independently address all questions arising from the archaeological research. Moreover, they are usually combined with other applications spanning from topographic mapping and geomorphometry of the landscape using LiDAR, DGPS units, robotic total stations and 3D laser scanners to aerial reconnaissance and satellite remote sensing (Fig. 14.14). Only the integration of the above-mentioned techniques with on-site excavations can provide a more holistic approach to the questions related to the interaction between human occupation and landscapes.

Fig. 14.14
figure 14figure 14

(a) Differential GPS being used to survey an archaeological site in NE South Australia. (b) Robotic total station being used for mapping a cemetery in Queensland, Australia. (c) Camera attached to a kite for acquiring aerial photos from the Neolithic site of Perdika 2 at Thessaly, Central Greece. (d) Satellite image of the urban plan of the ancient city of Ferai (Central Greece); pansharpened IHS transformation of the 15 June 2009 Quickbird image showing linear features associated with roads to the northern fields. (e) Same image as in (d); modified simple ratio (MSR) transformation of the 4 May 2010 GeoEye-1 image indicating the projection of the above features to the southern field (Sarris et al. 2015b)