ensors & Software loves to share customer stories in our newsletter! The details and descriptions are those of the author and Sensors & Software has not made any edits except for typographical errors. If you have a GPR topic of interest to share, please contact us and submit your suggestions.
In 2015, we acquired a NOGGIN® 500 SmartTow configuration, for archaeological exploration and teaching (Figure 1). Being new to GPR, to learn its capabilities, we conducted an intensive, 6-week GPR survey inside the Roman Fortress Castrum Cumidava in Rasnov, Romania (Figure 2).
The presence of a Roman fort was first documented by M.J. Ackner in 1856. The first excavations were carried out in 1939 but were never published. Excavations resumed in 1969 – 1974 but only the 1969 – 1970 results have been properly published. Since 2006, excavations have continued inside the castrum, but, so far, no results have been published beyond summary excavation reports.
Our survey covered the entire interior surface, except the areas heavily affected by current and past excavations Figure 2. XY grids of different sizes (to account for topographic imperatives) were set with line spacings of 0.5 m or 1 m. The total area covered by the GPR survey grids was roughly 120 x 160 meters.
As shown in Figure 3, the results are spectacular!!! We have a full and detailed overview of the layout of the castrum in its last phase. All the important, typical Roman streets have been identified:
1. Via Principalis (1), linking the southwest and the northeast gates,
2. Via Praetoria (2), perpendicular to the Via Principalis, (1) going from the central core of the castrum to the main (NW) gate,
3. Via Sagularis (3), around the interior of the castrum, along the fortification walls, and
4. Via Quintana (4), sub perpendicular to Via Praetoria (2), separating the North half of the castrum in two.
Via Principalis (1), divides the castrum in two major sectors: the North half is composed of wooden barracks, except for one single stone building in the western quadrant, fully excavated. The large number of GPR reflectors on both sides of the eastern half of Via Quintana (4) may suggest the presence of the various castrum shops, such as the smithy (blacksmith).
The South half of the castrum contains all the remaining stone buildings, with the core administrative building, the Principia (5), a large (approx. 17 x 20 m), complex structure with multiple rooms and four or five apses (semi-circular structures) on the South side.
On both sides, there are several adjacent buildings, all the way to the Via Sagularis (3), to the North. In front of the Principia (5), at the intersection of Via Principalis (1) and Via Praetoria (2), we have identified the unit’s altar (6). Behind it, several smaller, square structures of variable size, have been identified, possibly well(s), statue foundations, and/or other altars (6). To the West, we have confirmed the presence of a large (approx. 11 x 19 m) rectangular building, most likely a warehouse (horreum, 8).
The most important aspect of the GPR results is the information regarding the castrum evolution, exposing more of its “quirks”. The evidence available now from the archaeological excavations is mostly anecdotal, and tentatively identifies three separate phases. However, the GPR data seem to suggest the presence of four phases.
Phase 1: potentially from 101/2AD to around 117AD – it was made of wood and earthen fortifications, invisible to the GPR due to a combination of very weak signal, destruction, and reconstruction of later phases.
Phase 2: saw the construction of stone fortifications (9). However, at the intersection of the earliest iteration visible of Via Pretoria (2) and Via Principalis (1), the GPR data shows no traces of stone foundations of either the required altar or command structures, which indicates that all interior structures were still made from wood. This phase came to an end most likely with the first Marcomannic War, possibly around 170AD, with the partial or total destruction of the wooden structures.
Phase 3: construction resulted in an overall expansion of the fortified perimeter and the construction of the stone buildings in the South half of the castrum (10). The GPR data provides detailed mapping of almost all the stone structures in the camp. It also shows that Via Principalis (1) was moved 4 – 5 m to the SW and Via Pretoria (2) about 10 m to the West from the previous phase, which also meant that the NE and SW gates were moved, and, since Romans really liked straight roads and orthogonal layouts, by extension, the NW and SE gates were moved as well.
Phase 4: the changes from the third to the fourth (last) phase seem to indicate that the wooden structures inside the castrum, as well as the NW gate, were partially destroyed, possibly during the Germanic Invasions around 250AD. The GPR data shows that Via Principalis (1) was not straight any longer, since the gates at both ends were not aligned; the SE segment of the road remains unchanged, but the NW portion, starting from the stone altar (6) at the “center” of the castrum rotated at a 15o angle counterclockwise from its initial phase 3 orientation. Furthermore, the last iteration of Via Quintana (4) was not perpendicular to Via Pretoria (2) but parallel with the South segment of Via Principalis (1) indicating a large scale and perhaps complete restructuration of the area containing the wooden buildings, which – sadly – were not visible to the GPR.
In conclusion, the impressive detail of the structures of Castrum Cumidava provided by the NOGGIN® 500, at that time our newly acquired GPR system, did not disappoint!