Impact of earthquakes and its dependence on magnitude: testing the Greek seismicity

Strong earthquakes cause significant impact on both the built and natural environment. Impact databases are of fundamental importance for seismic risk assessment in a region. Such data include human and property losses as well as secondary effects including ground failures and tsunamis. The earthquake impact, EI, depends on many factors, one of the most important being the earthquake magnitude, M. To test the dependence of EI on M we selected the Greek seismicity which is the highest in the Mediterranean region with record of earthquakes since the antiquity. Although various descriptive and parametric earthquake catalogues as well as inventories of intensity observation points are available for Greece no database for EI has been organized so far. For a first time we organized a Greek Earthquake Impact Database (GEID) which covers the time interval from 1800 to 2019 and includes earthquake parameters and three main quantitative impact elements: building damage, fatalities and injuries. Data on tsunami impact are also included in the GEID. A long number of sources have been utilized, some of them remaining unknown so far in the seismological community. To select the most appropriate magnitude for each earthquake event occurring in the instrumental period of seismology, i.e. from 1900 onwards, we compared the catalogues produced by the ISC-GEM and by three academic institutions. After completeness testing and examination for magnitude homogeneity we performed magnitude closeness analysis and produced formulas for magnitude conversion from one catalogue to another. For the 19^th century earthquakes we again compared various catalogues, collected new data from documentary sources and compiled a new catalogue by re-calculating macroseismic magnitudes equivalent to Mw from intensity/M relations developed for Greek earthquakes of the instrumental period. We found that for single earthquake events the level of impact generally depends on magnitude but this is not valid for offshore events. However, the time distribution of the three impact elements over the period examined showed a relative decrease of the totally collapsed buildings which implied drastic decrease of the fatality rate but not of the injuries rate. This is attributed to the gradual improvement of the building construction particularly after the enforcement of antiseismic building codes in the country. Τhe first author was supported by the Hellenic Foundation for Research and Innovation (HFRI) and the General Secretariat for Research and Technology (GSRT), under the HFRI PhD Fellowship grant (GA. no. 490).