Last week, ground-penetrating radar (GPR, georadar) was used for the first time at Sagalassos. Excavation of the Domestic Area had almost reached the modern road to the west and the north. It became obvious that the complex continued northward, below the modern road. The task of the geophysical team was to provide a map of the buried structures in the unexcavated area directly northwest of it and of the ongoing excavation in courtyard XLV and room XLVI (see Domestic Area, this week). Under such conditions, geophysical exploration is strictly limited to GPR as other methods would be much less efficient. The aim of this GPR prospection in the city center was also to evaluate its potential in urban conditions. We need to know the potential and the possible limitations of this technique before designing future geophysical prospections in the city center. The main obstacles are varied and uneven surfaces with stones and lots of collapsed material below the surface, as well as unknown depths and levels of preservation of architectural remains. Because of the ruined state of the remains, some time was spent in the field before an appropriate compromise between the depth of penetration and the noise ratio signal was reached. In these circumstances, GPR proved to be a very efficient technique. A series of time slices was generated to visualize the amplitudes of georadar echoes at several depths. Architectural remains represented by high amplitude rebounds were clearly depicted, and we could conclude that these remains represented the continuation and end of the palatial complex to the northwest. It can be clearly seen that the western wall is very thick (more than 1 m) and that the northern one is much thinner and associated with some other, even thinner and badly preserved walls (see illustration s.v. Domestic Area, July 24-28, 2005).

In the Potters' Quarter, the aim of the GPR prospection was to provide more detailed information about specific targets that cannot be interpreted from magnetograms, such as water channels and the whole water-supply system of Sagalassos. As a test for detecting water channels by GPR, we selected an area adjacent to an excavated water channel made of stone in the northern part of the Potters' Quarter. On the georadar profile, which represents a vertical cross section through the ground, the water channel could easily to be distinguished. Based on these results, one can conclude that water channels can be traced at least in areas where they are not too deep or not covered by thick layers of ruins.

The potential of georadar scanings was also tested in the northeast corner of the gymnasium east of the theater (see Gymnasium, 2004). While magnetometry had produced a plan with some rooms, the GPR scan produced much more detailed plan views at certain depths. We believe, that in such conditions it is possible to recognize even different phases of the building history. Another test produced plan views of a workshop with an internal division into smaller spaces and two perpendicular streets with a crossroad in the near vicinity. In the Potters' Quarter magnetograms and radargrams are complementary to one another in the way that magnetometry produces a much better visibility for kilns, whereas architectural remains are clearer on radargrams.