A great amount of global seismicity occurs in regions with very limited station coverage and significant error in estimates for seismic traveltimes can lead to considerable event mislocation and a very blurred image of seismicity. We present case studies in which modern probabilistic event location algorithms provide dramatically improved images of the distribution of seismic events. So-called precision seismology has vastly improved relative event location estimates over short distances for which waveform similarity is observed. For example, the declared nuclear tests in North Korea can be located relative to each other to within an accuracy of less than hundreds of meters. However, we can demonstrate that the outgoing wavefield is more complicated than is assumed in most double-difference type location estimates. Precision seismology will improve greatly if it can encompass the progress made in the context-based multiple event location algorithms and account for deviations from the applied models resulting from features on smaller length scales than we are currently able to model.
|Last updated||Key Points|