Divergent Responses of Cyclades Islands to a Major Aegean Earthquake: A geophysical perspective from Amorgos Island

On 9 July 1956 at 05:11 local time, a M_w7.5 earthquake struck the southeastern Cyclades (Greece), followed thirteen minutes later by a second event of magnitude M_w7.1–7.2. The combined effects of these two earthquakes, whose epicentres were located between the islands of Amorgos and Santorini, resulted in 54 fatalities and caused severe damage on Santorini, leading the Greek authorities to declare the island a “large-scale local disaster.” In contrast, the situation on Amorgos is less well documented. Although reported macroseismic intensities (EMS98) on the island range between VI and IX, recent field surveys and interviews conducted with local inhabitants, within the framework of the ANR-Amorgos project, suggest that the actual impact was limited. This discrepancy raises a key question: how can such markedly different seismic responses be explained between Amorgos and Santorini, given that Amorgos lies less than 30 km from the rupture zone? While several geophysical studies have focused on Santorini to characterise its geological structure and site response, no comparable investigations had previously been carried out on Amorgos. To address this gap, we conducted a large geophysical survey combining HVSR, MAM (Microtremor Array Measurements) and MASW (Multichannel Analysis of Surface Waves). The aim was to characterise site response in the island’s main inhabited areas and to compare these results with the unexpectedly high macroseismic intensities reported for the 1956 earthquakes. A comparative analysis with existing data from Santorini was also performed to identify the factors responsible for the contrasting seismic responses observed between the two islands.

Our results provide a geophysical site characterisation of the main urbanised sectors of Amorgos (Aegiali, Katapola, Chora and Arkesini). No significant site amplification effects have been identified to date, neither within sedimentary basins nor along topographic highs. Moreover, the fundamental ground frequencies obtained are significantly lower than the relatively high resonance frequencies typical of traditional Cycladic buildings (>10 Hz), suggesting limited double resonance effects. At this stage, the available geophysical data do not account for the strong spatial variability of macroseismic intensities reported on Amorgos Island and call into question the reliability of these historical observations. By contrast, the presence of thick volcanic formations (first and second explosive cycles: 2-300 kyr) combined with steep topography appears to be the primary factor explaining the much more severe damage observed on Santorini during the 1956 seismic sequence.