Publication Date:
2020-10-08
Description:
Summary A local seismic magnitude scale, ML$^{sf P}$, has been developed for the United Kingdom (UK) using automated measurements of 8902 half peak-to-peak vertical component seismic P-wave displacement amplitudes from 630 earthquakes. The measurement time window increases with source-to-receiver range such that ML$^{sf P}$ is sensitive to the dominant phase within the P-wavetrain at a given distance. To avoid contamination due to low-frequency noise the P-wave amplitude measurements are made in the 1.5 to 30 Hz passband. A least-squares inversion was undertaken to estimate source size, distance and station effects. The distance effect values suggest that P-wave amplitude attenuation across the UK is low when compared to other tectonically stable regions. The station effects are broadly consistent with UK geology, with signal amplification observed within the sediments towards the south-east of the country. ML$^{sf P}$ has been tied to the UK local magnitude scale routinely estimated by the British Geological Survey (determined using S-waves, and here denoted ML$^{sf BGS}$). For earthquakes with ML$^{sf BGS}$〉3, ML$^{sf P}$ exhibits a closer correspondence to the moment magnitude than ML$^{sf BGS}$ (i.e., ML$^{sf P}$≈M$_{sf W}$). It is tentatively suggested that this reduction in bias is caused by theP-wave scale being less affected by along-path attenuation. The difference with respect to physical source scaling helps explain the divergence of the ML$^{sf BGS}$ and ML$^{sf P}$ scales at ML〉3. ML$^{sf P}$ allows a robust estimate of event size to be made for small events which predominantly generateP-waves, e.g., near-surface explosions. ML$^{sf P}$ values have been calculated for 239 explosive events, mostly mining blasts and munitions disposal. Although there is significant scatter, explosive events exhibit elevated ML$^{sf P}$ values compared to ML$^{sf BGS}$, consistent with explosions generating proportionally more compressional wave energy than earthquakes. For example, 33 explosions at sea exhibit a median ML$^{sf P}$-ML$^{sf BGS}$ value of 0.50 magnitude units. Despite its sensitivity toP-wave amplitude, ML$^{sf P}$ is not a more consistent estimator of explosive source size than ML$^{sf BGS}$; the magnitude residuals (station estimate - event estimate) are slightly less for ML$^{sf BGS}$ compared to ML$^{sf P}$. This is primarily due to variability of theP-wave amplitudes that cannot be explained by a 1D distance correction. ML$^{sf P}$ should be considered as an additional tool for characterizing small seismic events within the UK.
Print ISSN:
0956-540X
Electronic ISSN:
1365-246X
Topics:
Geosciences
Permalink