Non-destructive geophysical survey methods

The 2015 and 2016 excavation campaigns were each preceded by a geophysical survey led by Dr. Apostolos Sarris, director of the Laboratory of Geophysical- Satellite Remote Sensing & Archaeo-environment (Foundation for Research & Technology, Hellas). The approach adopted by the geophysical team favoured high resolution over wide coverage. Specific areas of limited extent were surveyed intensively and using different methods, so as to provide archaeologists with a detailed map of these sectors and in this way inform the excavation strategy. The goal of the 2015 geophysical survey was to locate and map architectural remains on the south-eastern terrace of the hill, where a potential court-centred building (Building F) had been identified in 2011 (Area C) (Figure 1). In 2016, the investigations focused on Area A, i.e. to the north of Building CD, and in Area B, i.e. between Building CD and Building F.



Figure 1. Grids surveyed in 2015 (Area C) and 2016 (Areas A and B) (A. Sarris)

Methods, instruments and data collection strategy

Geophysical methods measure the physical properties of soils on or below the surface. In 2015, based on the geomorphological characteristics of the hill and the nature and depth of the expected archaeological targets, a decision was made to employ three different geophysical techniques in Area C: ground penetrating radar (GPR), electrical resistance and magnetic gradiometry. Electromagnetic methods like GPR are ideal for obtaining information on both conductivity and magnetic susceptibility of soils when large penetration is needed. Soil resistance techniques are best suited for features that contrast with the surrounding soils in porosity, density and water content, like walls and ditches. Magnetic methods, on the other hand, are particularly suitable to identify features that contrast with surrounding soils due to concentrations of magnetic minerals, e.g. burned features, habitation units and ditches filled with organic material. The instruments implemented at Sissi are illustrated in figure 2: 1) the Bartington fluxgate gradiometer G601, 2) the Noggin Plus GPR with 250 MHz and 500 MHz antennas, and 3) the GeoScan Research resistivity meter RM85. The GPR survey was carried out in zig-zag mode, moving along parallel transects 50 cm apart, with a sampling interval set at ca. 2.5 cm along the transects. The measurements were taken twice, first along the x-axis (i.e. in a north-south direction) and then along the y-axis (i.e. in an east-west direction). This strategy doubled the surveying and processing time, but it also substantially increased the resolution of the data. A sampling interval of 50 cm along the x-axis and 12.5 cm along the y-axis was chosen for magnetic measurements, whereas soil resistance data were collected at an interval of 1 m by 50 cm.


GPR proved the most effective geophysical technique in Sissi in 2015. In 2016, it was therefore decided to decrease the sampling interval to 25 cm by ca. 2.5 cm, in both Areas A and B. In Area A, soil resistance data were collected with a sampling interval of 50 cm along the x- and y- axes. However, rather poor results were obtained, probably due to the dry conditions of the subsoil, leading us not to employ electrical resistance technique in Area B. Magnetometry, which was not particularly successful in 2015, was not used at all in 2016.



Figure 2. Geophysical instruments employed at Sissi.


Clear GPR data were obtained in Area A (Figure 3), suggesting the existence of a series of small rooms to the north of Building CD. The orientation of these rooms is similar to that of the compartments excavated in Building CD (e.g. Rooms 4.5, 4.8, 4.9 and 4.10). Room A1 is ca. 2 by 2 m large. To the north, A2 and A4 have a long axis of ca. 4.5 m, but both seem to be subdivided by inner walls. Still further north, GPR signals indicate a much larger room (A3), ca. 8 by 5.30 m in size. There is evidence for three more compartments (A5, A6 and A7) with internal divisions between A3 and Rooms 4.8 and 4.5. A thick wall to the east of A6 and A7 separates these rooms from another unit comprised of A8, A9 and A10. Room A9 seems to be the largest of them, measuring ca. 2.8 by 3.8 m. To the east of A9, the elongated feature A11 may be a road segment.



Figure 3. GPR results from Area A. Top left and top right: processed GPR data. Bottom: Diagrammatic interpretation of GPR reflectors (A. Sarris)

Fuzzier results were obtained in Area B, although several anomalies suggest the existence of built features (Figure 4). Structures B1, B2, B3, B4 and B5 are the most clearly defined. Their walls are oriented north-south and east-west. Rooms B1 and B2 seem to belong to the same complex, ca. 5 m long and by 2.5 m wide.  Further to the north, rooms B3 and B4 both measure ca. 2.3 by 2.5 m. Building B5, which can be traced for a distance of ca. 7 m, consists of two or even three rooms, all ca. 2 m wide. There is also evidence for additional built features to the west and to the north (B7, B8 and B9).



Figure 4. GPR results from Area B. Top left and top right: Processed GPR data. Bottom: Diagrammatic interpretation of GPR reflectors (A. Sarris)

In the northern part of Area C, information coming mainly from GPR and partially from magnetometry points to a number of compartments with an orientation roughly similar to the rooms excavated to the south-east (Figure 5). The most evident feature is the two-room structure C1, ca. 5 by 3.5 m. More linear anomalies (C2) suggest the continuation of the habitation quarters to the north and to the west. There is also evidence of multiple walls delimiting several compartments in the vicinity of C3 and C5. However, the most striking feature identified by means of geophysics is a large area, ca. 30 m long and maximum 14 m wide, almost completely devoid of anomalies or reflectors – with the exception of C4 and C7. This large area corresponds to a court, the north and east edges of which were excavated in 2011. The uniform signature of the court is clear in GPR and soil resistance data, both of which show an absence of internal built features. The outline of the central court is clearly defined to the west and to the south by linear segments that meet at a right angle to the south-west. A similar linear segment is also visible to the south-east of the court, near Room C9. This segment probably extends north up to the façade of the rooms excavated to the east of the court. Finally, GPR and magnetic data allowed the identification of an independent rectangular building (C8) in the south-western part of Area C. The magnetic data also indicate a couple of extreme magnetic features (thermal targets suggesting residues of burning) in the same region.



Figure 5. Results of the geophysical survey in Area C. Top left and top right: processed GPR data. Bottom: Diagrammatic interpretation of all geophysical features (A. Sarris)


Apostolos Sarris, Meropi Manataki & Sylviane Déderix