Large-amplitude ground-motion recordings and their interpretations

Empirical recordings of ground motions featuring large amplitudes of acceleration or velocity play a key role in defining the maximum levels of ground motion required for the design of engineering projects, in view of their potentially very destructive nature. They also provide valuable insights as to the nature of the tails of the ground-motion distribution. The present paper reviews the interpretations put forward for a dataset of recordings selected for their large value of PGA (⩾1 g) or PGV (⩾100 cm/s), and attempts to classify causative physical processes according to a uniform nomenclature. The aim of the study is to identify to what extent large-amplitude ground motions are predictable, by separating scenarios that could generally be encountered and foreseen using current knowledge of the processes involved in earthquake ground-motion generation and propagation, such as particular features of the recording site, from occurrences that are related to unforeseeable characteristics of the source and path.