Induced earthquake hazards: Forecasting seismicity rates using a hydromechanical earthquake nucleation model and saltwater disposal data
Jack Norbeck, USGS, ESC, Menlo Park
Wednesday, August 16, 2017 at 10:30 AM
- Building 3, Rambo Auditorium
- Ole Kaven
The earthquake activity in Oklahoma and Kansas that began in 2008 reflects the most widespread instance of induced seismicity observed to date. Given the continued high rates of seismicity in this area, including three M5+ earthquakes in 2016, it is imperative to develop methods that forecast future earthquake hazard accurately. In this work, we develop a method that combines a fluid pressurization model with rate-and-station friction theory to forecast rates of induced seismicity. Using injection well data from 893 saltwater disposal wells completed in the Arbuckle aquifer, we established a detailed description of the temporal and spatial variability in stressing conditions over the 21.5-year period from January 1995 through June 2017. Seismicity rate forecasts generated based on these stressing conditions replicate many salient features of seismicity in Oklahoma and southern Kansas including the onset of the sequence, the timing of the peak seismicity rate, and the reduction in seismicity following decreased disposal activity. Comparing our hydromechanical earthquake rate model to observational and empirical models that are commonly used to forecast seismicity rates, we find that our model outperforms them for 1-year forecast durations for the years 2008 through 2016.