High-detail damage pattern in towns hit by earthquakes of the past: An approach to evaluate the reliability of the historical sources

Abstract

Historical data relating to damage in urban areas are useful to evaluate and mitigate seismic risk. However, documentary data need to be considered in relation to their quality. The paper describes a mathematical approach to assess reliability of historical sources referring to damage effects caused by earthquakes in the past. The method applied is based on Rough Set Theory. Its application has been explained by analyzing a huge amount of archive data reported by documents concerning earthquake effects in eight towns of Basilicata (southern Italy), heavily damaged by the 1930 Irpinia earthquake. Mathematical analysis has been applied to evaluate whether technical reports citing the effects on buildings were affected by voluntary or involuntary “manipulation” by compilers. This has been possible by comparing damage descriptions reported in technical reports to damage inferred by the analysis of independent technical-economic-administrative data. Results seem to suggest that the sentences on damage were subjected only to limited and involuntary “falsification” that, however, caused negligible changes in the spatial distribution of damage.