The earthquake of 250 a.d. in Augusta Raurica, A real event with a 3D site-effect?

The Roman city Augusta Raurica is located East of Basel, Switzerland. One important topic of the city’s history concerns the hypothesis of an earthquake striking the city in the middle of the third century a.d. This idea had been formulated according to archaeological features and findings, but had not been tested so far. A selection of the archaeological features were reviewed and dated in order to test the hypothesis of a single event. However, archaeological investigations do not draw a conclusive picture; it could not be proven that all features of possible destruction date to the same event. Detailed seismological investigations were performed. These included geological and geotechnical mapping of the unconsolidated sediments. Important parameters such as the thickness and composition of the unconsolidated sediments, the terrain topography and the topography of the bedrock surface were mapped. Ambient vibration H/V measurements provided the fundamental frequency of resonance for the unconsolidated sediments. The velocity of shear waves traveling through sediments is the controlling parameter for amplification of seismic waves. This material property is estimated using the relation between the ellipticity of the fundamental mode Rayleigh wave and the H/V curve. From all information we compiled a three-dimensional model of the surface geology. This model is used to simulate earthquake ground motion and amplification effects in the city, and to map the variability of the amplification. In the part of the city where possible earthquake damage was recognized, amplification occurs in the frequency band of building resonance (2–8 Hz). In the other part of the city amplification occurs much above the building’s resonance. From 1D modelling we estimate a difference in spectral amplification of about a factor of 2.5 to 3 between the two parts of the city. This corresponds approximately to a difference in macroseismic intensity of one unit. 3D modelling showed a large variability of ground motion within very close distance in the part of the city where possible earthquake damage was recognized. The maximum amplification reaches values up to a factor of nine, which is due to 3D effects and the choice of using vertically incident waves. Finally, all paleoseismological findings in the area of Basel were reviewed in order to find indications of a large event in the time-period of interest. Paleoseismological findings provide no hints to a large earthquake in the third century. If we assume that an earthquake caused at least part of the identified damage in Augusta Raurica, we have to assign to this event a magnitude Mw of about 6.0 or even lower, that is much smaller than the value of 6.9 that is actually in the Swiss earthquake catalogue. The earthquake source of this event must then be very close to the site of Augusta Raurica and a strong site-effect occurred in one part of the city.