Forecasting Aftershock Activity: 5. Estimating the Duration of a Hazardous Period

Abstract

Continuing the series of publications on aftershock hazard assessment, we consider the problem of estimating the time interval after a strong earthquake that is prone to aftershocks which may pose an independent hazard. The distribution model of this quantity, which depends on three parameters of the Omori–Utsu law, is constructed. With the appropriate averaged parameter estimates, the model fairly closely fits the real (empirical) distributions of this quantity on the global and regional scale. A key parameter in the model is the expected number of aftershocks of a given magnitude. This number broadly varies from earthquake-to-earthquake, which determines the wide confidence variant of the estimates based on the averaged parameters. Therefore, for forecasting the duration of the hazardous aftershock-prone period, we propose to use two variants of the estimates. The first variant is based only on the averaged parameter estimates for the region under study and on the value of the magnitude of the earthquake. This variant is applicable immediately after a strong earthquake. The second variant employs information about the aftershocks that occurred during the first few hours after an earthquake, which improves the forecast considerably.