CRC 1266 - Scales of Transformation

Phase 1 - Research activities 2016-2020

G2: Geophysical prospecting, classification and validation of settlement remains in changing environments

Principal Investigators: Prof. Dr. Wolfgang Rabbel, Dr. Dennis Wilken, Prof. Dr. Thomas Meier
Staff: Clemens Mohr, Diana Panning, Natalie Pickartz, Erica Corradini


Research agenda

Transformations of socio-environmental interactions are often associated with changes in the construction of settlements and/or ritual sites. The major objectives of sub-project G2 are to apply and refine existing geophysical investigation methods to the key locations of the archaeological sub-projects of the CRC.

Sketch working areas
Fig. 1. Sketch illustrating how the different working areas of G2 interact. The two different geometrical forms in A represent two classes of archaeological objects that will be investigated for their shape (A) and then the interior buildup of key targets in terms of geophysical soil parameter distribution will be determined (B). Coring and excavation related geophysics, in other words in-situ estimation of these soil parameters and their corresponding soil types will be determined in D. Resolution of B will be enhanced by feedback resulting in a 3D subsurface model of these objects. The derived data inversion and structural knowledge of the key targets will then be applied to non-excavated targets (E).

Data examples
Fig. 2. Data examples corresponding to the different working steps and approaches sketched in Figure 1. (A): example of a larger scale magnetic prospection; the inset shows a selected pithouse which was further investigated in B-D. (B): example of multiple geophysical methods (GPR, Seismics, Geoelectrics) applied to the pithouse. (C): possible model of the pithouse (left) derived from geophysics (B), coring and excavation (D) (consisting of structural information and physical soil parameters) and an example of a modeled magnetic dataset based on that model (right).


Form 2017 to 2019 geophysical measurements were performed at the paleolithic site of Horsens, the Mesolithic site of Duvensee, the neolithic sites of Vráble and its larger surrounding, Stolniceni, Maidanetske, and the Bronze Age site of Bornhöved. Exemplary results from phase I are described below; an extended description can be found in section Research activities.

We developed a new geophysical methodology for determining the spatial distribution of magnetized remains (daub) of houses and their masses from magnetic maps at Neolithic mega-site of Maidanetske (Ukraine). It enables to determine the find density and the internal patterns of burned houses, to test hypotheses about the burning processes of the houses, and a new way of house classification.

Systematic interlinking of geophysics, drillings and excavations, was used for developing methodologies that were successfully applied for determining a 3D model of the stratigraphy and the siltation and landscape evolution of the Mesolithic site of the Duvensee bog (Northern Germany), and for extending 2D archaeological documentation from plana and trench-type excavations into 3D models as demonstrated for Neolithic house trenches at the Vrable location (Slovakia).

Research activities

Form 2017 to 2019 geophysical measurements were performed at the paleolithic site of Horsens, the mesolithic site of Duvensee, the neolithic sites of Vráble and its larger surrounding, Stolniceni, Maidanetske, and the bronce age site of Bornhöved. Besides standard areal geophysical prospecting at the earth surface we performed point measurements in archaeological excavations and trenches, in boreholes and at samples from corings. This data served for calibrating and ground-truthing geophysical surface measurements, for defining proxys for lithology at different scales, and for establishing transfer functions relating geophysical and non-geophysical soil properties.

2017 - Neolithic site of Maidanetske (Ukraine, in cooperation with subproject D1)

We developed a new geophysical methodology for determining the spatial distribution of magnetized remains, basically daub, of Neolithic houses and their masses from magnetic maps. The calibration of the method is based on distribution maps of different find categories such as daub, ceramics and stone artefacts, which are quantified in terms of mass per square meter and magnetic susceptibility. This information was used to derive spatial magnetic models for each find category. The corresponding magnetic fields per find category were computed numerically and matched with the archaeological data. The new joint magnetic and archaeological method allows estimating the extent and find density for the whole extraordinary large archaeological site.

Ausgrabung Maidanetske
Fig. 1. Excavation and planum based geophysical exploration in Maidanetske 2017.


2017 - Neolithic site of Stolniceni (Moldavia, in cooperation with subproject D1)

Susceptibility measurements in a dense raster of one excavation section demonstrate the potential of documenting archaeological structure by high-resolution physical measurements, additionally to traditional archaeological documentations. In addition the resolution and depth penetration of GPR-measurements was analysed. The Stolniceni data enable a joint inversion of GPR and ERT data to be performed in the near future, providing information about the depth and geometry of the archaeological structures imaged in the magnetic map.

2017 & 2018 - Neolithic site of Vrable and its surroundings (Slovakia, in cooperation with subproject C2)

The field measurements in Vrable were concentrated on systematic drilling of exemplary magnetic anomalies and determining a comprehensive data base of susceptibility-depth functions and lithology. Besides statistical analysis, complete susceptibility transects were acquired allowing to reconstruct magnetic subsurface models and to determine synthetic magnetic data by numerical modelling. The comparison of synthetic and measured data showed that induced magnetization, as derived from the susceptibility distribution, could explain only up to 30 % of the measured magnetic field strengths. This implies that remanent (“permanent”) magnetization of the subsurface material plays an important role.

Comparative measurements at the Vrable site before and after excavation demonstrated the strong influence ploughing has on GPR and EMI measurements and thus imply the preferable application of geophysical methods directly on the excavation planum. Using the example of a house pit we could show that combined EMI and  GPR measurements are capable to extend the documentation of the excavated area from 2D to full 3D beneath the planum. The measurements showed the non-uniform depth extent and filling of house pits, and they enable an improved planing of location and required depths of excavation trenches (Pickartz et al., 2019c).

Mira Vrable
Fig. 2. Usage of the MIRA multichannel GPR system in Vrable.

2018 & 2019 - Bronze-age site of Mang de Bargen near Bornhöved (Germany, in cooperation with subproject D3)

Large area excavations of the D3 group enabled a substantial ground-truthing of magnetic anomalies and objects located by areal GPR measurements beforehand. This comparison revealed the fundamental problems of the interpretation of geophysical data in this glacially formed soil environment, which are caused by high similarities of geologic and anthropogenic structure (which could sometimes not even be distinguished despite excavation). This initiated an ongoing study on geophysical attributes helping to interpret the features and a collection of indistinguishable situations to identify possible pitfalls. Newly acquired GPR data bear a wealth of highly resolved structure enabling, inter alia, the identification of construction details of grave mounds and mass balancing of trenches and grave mounds.

GPR Bornhöved
Fig. 3. GPR measurements on the planum during excavations in Bornhöved.

2017, 2018 & 2019 - Mesolithic site of Duvensee (Germany in cooperation with subproject B2)

Based on two extensive GPR surveys at the Duvensee site we could recognize the location, depth extent and size of the former, now silted-up, islands, lakes and land bridges of the area. Especially, the islands where Mesolithic camps were located according to archaeological finds could be outlined. The initial coarse gridded GPR survey was later on  refined and extended with new instrumentation (Mala MIRA System). In this way details of previous shorelines could be reconstructed and possible locations of Mesolithic settlement places could be narrowed. In particular, it could be verified that a hazelnut roasting place, which had been found by chance, is underlain by a soil interface locatable by GPR. Besides the described prospection work progress was made in incorporating lithological information and vertical electrical soundings from drillings as constraints for electric resistivity tomography in order to improve stratigraphic interpretation. Based on this integrated approach a 3D model of the temporal landscape evolution could be derived that fits well with spatial distribution and ages of the mesolithic finds.

Fig. 4. Geoelectric, GPR, and downhole measurements in Duvensee.

2018 & 2019 - Palaeolithic site of Horsens (Denmark, in cooperation with subproject B1)

At the Horsens site intensive GPR and ERT surveys had been performed in an attempt to detect and outline kettle holes at depth. Kettle holes are known as palaeolithic find spots. Based on these measurements two promising sites were further investigated by EMI and trenching. The excavated strata show a close correspondence to the stratigraphy found by GPR. Also EMI was successful in outlining the near-surface contour of the suspect kettle hole.


Alongside the geophysical measurements at the different sites we performed methodological work:

  •  Comparative interpretation of geophysical surveys to reconstruct the environment of archaeological sites and to detect archaeological features.

Duvensee Moor
Fig. 5. Top – Comparison between GPR and corings and identification of the main facies of the Duvensee bog. Based on the, a reconstruction of the Duvensee bog with a hypothetical regression of the water level and occuptaion of the area was constructed (below).

  •  Development of approaches for performing and interpreting down-hole measurements of electrical resistivity and dielectric permittivity.

Fig. 6. Application and results of geoelectric downhole measurements in Duvensee.

  •  Numerical modelling of synthetic geophysical data for methodical tests and for supporting field data interpretation.
  •  Comparison of surface and down-hole measurements and development of algorithms for an integrated interpretation.

A summary can be found in a digital contribution to the AGU Fall meeting 2019:

  •  Comparison of the accuracy of four different ways of determining magnetic susceptibility-depth-curves in situ and in the lab. It turned out that the most reliable approach is to measure susceptibilities in situ (down-hole or at exposures). Guidelines for upcoming fieldwork have been developed.
  • Development of modelling and inversion algorithms of magnetic data.
  • Determination of remanent magnetization for magnetic field data and in-situ susceptibility measurements.
  • Development of an approach to determine the masses of burnt Neolithic houses from magnetic measurements.

Magnetik Maidanetske
Fig. 7. Magnetic map of the Maidanetske site (Ucraine, after Rassmann et al. (2014)) with different discovered house types enlarged. The inset shows the estimated total mass per area versus total mass of houses in Maidanetske. The curves show two different sets of house types as well as megastructures.

A summary can be found in a digital contribution to the AGU Fall meeting 2019.

The results have been presented at different scientific conferences like the International Conference on Arhcaeological Prospection in Sligo, Ireland (2019) and the Annual Fall Meeting of the American geophysical Union in San Francisco, USA (2019).

Präsentation Duvensee
Fig. 8. Presentation of the Duvensee results at the AGU 2019 Fall Meeting.






New Publications