Unearthing Aftershocks: Physical Simulations Statistical Models, and New Observations
DESCRIPTION: Earthquakes occur not as single isolated events, but as clusters of events. In the aftermath of an earthquake, many smaller ones happen, called aftershocks, at a rate which decreases with the time following an event. Occasionally one of the aftershocks is even larger than the mainshock that triggered it. These aftershocks are important to understand, both for the insights they give us into earthquake processes, but also as a practical measure for society, as understanding them better aids in preparing for and recovery from the aftermath of a large earthquake. One aim of this work is to develop scientific knowledge which can contribute to a new effort to develop Operational Earthquake Forecast capabilities, which will provide automated tools for emergency managers to receive real-time updated forecasts using aftershock clustering properties to better quantify hazards, for use in response and recovery and planning, as well as outreach and communication to the broader public. The investigators on this project are part of a team working to develop that system. Aftershocks of earthquakes are particularly interesting phenomena in that they have been shown to have a wide variety of statistical regularities, and yet remain, at the core, not understood, in the sense that a physical model has yet to be developed which can reproduce all the statistical regularities. In the absence of a physical model, a number of statistical models have been built to develop synthetic catalogs and forecasting methods, among other uses. But what statistical ingredients to include remain uncertain. This proposal seeks to develop a better physical understanding of aftershocks from a variety of approaches, and apply this understanding in a number of useful ways. Three specific elements are proposed. One is a new physical model, which treats a fault as a multi-stranded rough set of surfaces, and is capable of reproducing the spatio-temporal features and productivity statistics. The second one is a new type of statistical measurement, which uses the abundance of aftershocks close to the rupture area to better resolve the spatial relationships among aftershocks and the mainshock. The third one is a new set of statistical models to integrate with time dependent hazard models, enabling generalizations to incorporate the new non-separable magnitude-spatial kernels indicated by the observations. Success in any of these elements could further help in quantifying earthquake hazards, through improved physical understandings of ingredients in empirically based ground motion relations used to estimate and design for engineering demands, and scaling and clustering behavior in operational earthquake forecasting, a tool of significant societal use in response and planning in the aftermath of a large earthquake. The project would aid in the training and mentoring of an early career scientist. The project would also support the participation of the PI in organizations developing operational earthquake forecasting capability, WGCEP (Working Group on California Earthquake Probabilities), and their application, NEPEC (National Earthquake Prediction Evaluation Council).