Further information on the macroseismic field in the Balkan area

Abstract

Papazachos and Papaioannou (1997) (called PP97 hereinafter) studied the macroseismic field in the Balkan area (Greece, Albania, former Yugoslavia, Bulgaria and western Turkey) with the purpose of deriving attenuation and scaling relations useful for seismic hazard assessment and study of historical earthquakes. In his comment, Trifunac suggests that our analysis might exhibit certain bias for all countries except Greece due to problems mainly associated with the database (completeness, etc.), conversion of local intensity scales used in the Balkan countries, as well as to the local variations of the attenuation relation due to the variation of the geotectonic environment in this area. Specifically, his most important comments can be summarized as follows: a) The large participation of Greek data probably biased the scaling relations proposed in the study. b) The conversion relations used between local macroseismic scales are less accurate than their proposed such relations. c) The variation of attenuation (geometrical and anelastic) in different regions of the study area is important and local relations (instead of the proposed single relation) should be determined for seismic hazard assessment. In the following, we study in detail each of these possible bias sources. Additional work on the macroseismic field of the Balkan area shows that none of the previously described factors, suggested by Trifunac, introduces bias in the results presented by PP97. Specifically, it is shown that the database used by PP97 fulfills the basic requirements for a reliable determinations of attenuation and scaling relations proper for seismic hazard assessment in all five countries of this area. Evidence is presented that no strong geographical variation of the attenuation of macroseismic intensities of shallow earthquakes is observed. Relations between local version of intensity scales suggested by Shebalin et al. (1974) are shown to be reliable. Finally, it is demonstrated that national practices for estimation of macroseismic intensities may affect the results of seismic hazard assessment but proper formulation can be applied (PP97) which allows to take into account such differences in national practices. This formulation allows also to introduce and correct for anisotropic radiation at the seismic source as well as the incorporation of site effects.