Prediction of Coseismic Rupture Locations from Fault Maps

Coseismic fault displacement is a localized source of hazard following surface-rupturing earthquakes. Here, we exploit the usefulness of geological maps and fault traces contained therein, to predict the location of coseismic ruptures (CORs). We analyze five earthquakes that occurred inItaly and the UnitedStates, measuring (1) the amount of CORs that occurred along previously mapped faults and (2) the amount of already mapped faults that actually moved during a strong earthquake. These quantities are 32% ± 30% and 20% ± 11%. The values are highly variable depending on the investigated earthquake and are influenced by the distance from the principal fault, the structural ranking, and the quality of geological maps. Ruptures structurally connected and close to the principal fault are more easily mapped. Weinvestigate whether geological maps could be exploited in the context of hazard assessment, and we suggest that specific regressions which consider available fault maps could be applied to compute the probability of occurrence of distributed faulting. Moreover, we perform a preliminary binomial logistic classification to identify a distance threshold from the principal fault to infer possible fault reactivation. A reliable geological map is a useful tool for hazard and fault capability assessment, land planning and resource management, and prompting for the need for proper resource allocation. Some critical aspects that need to be addressed in the future include how to properly include geological knowledge and how to treat triggered ruptures in modeling approaches.