Seismicity, Critical States of: From Models to Practical Seismic Hazard Estimates Space

Abstract

The most fundamental question in earthquake science is whether earthquake prediction is possible. Related issues include the following: Can a prediction of earthquakes solely based on the emergence of seismicity patterns be reliable? In other words, is there a single or several “magic” parameters, which become anomalous prior to a large earthquake? Are pure observational methods without specific physical understanding, like the pattern recognition approach of Keilis–Borok and co-workers [41], sufficient? Taking into account that earthquakes are monitored continuously only for about 100 years and the best available data sets (“earthquake catalogs”) cover only a few decades, it seems questionable to forecast earthquakes solely on the basis of observed seismicity patterns. This is because large earthquakes have recurrence periods of decades to centuries; consequently, data sets for most regions include less than ten large events making a reliable statistical testing questionable. In the studies discussed here, the goal is not to forecast individual earthquakes. Instead, we aim at developing a combined approach based on numerical modeling and data analysis in order to understand seismicity and the emergence of patterns in the occurrence of earthquakes. The discussion and interpretation of seismcity in terms of statistical physics leads to the concept of “critical states”, i. e. states in the seismic cycle with an increased probability for abrupt changes involving large earthquakes. A more general goal of this work is to provide perspectives for the understanding of the relevant mechanisms and to give outlines for developments related to time-dependent seismic hazard.