ABSTRACT

Rapid and accurate estimation of event magnitudes is important for real‐time monitoring, especially in areas where a traffic‐light protocol for induced seismicity has been established. In some instances, event magnitudes that have been reported by different groups exhibit an undesirably high degree of scatter. This may be caused by various factors, including the use of different station configurations, mixed sensor hardware with varying instrumental response, waveform clipping, or lack of magnitude standardization. In this study, we use a recently acquired hydraulic‐fracture‐monitoring case study from western Canada to evaluate the use of a coda‐duration method to estimate magnitudes for moderate and small events in a range 0.5Mw3.6. The dataset was acquired using a mixture of sensors, including shallow‐buried broadband seismometers and three‐component 10‐Hz geophones cemented at 27 m depth. Although local calibration is required, the coda‐magnitude method has certain advantages over traditional methods for rapid magnitude estimation, including relative insensitivity to waveform clipping, uncertain attenuation corrections, or both. By calibrating measured coda duration with independent estimates of Mw derived from low‐frequency spectral amplitude values, we obtained a trilinear representation for coda magnitude that is similar to previously published scales for other areas. Measurements of radiated S‐wave energy flux, obtained as a by‐product of the coda calculations, are consistent with previously reported energy relations and support an interpretation that underlying physical rupture processes are scale‐independent over the size range sampled by our dataset.

You do not currently have access to this article.