In both the single-station data shown in Fig. 4 and in the array data in Fig. 5 a single depth value CCT244747 unable to precisely match the observed depth phase delay times at all stations, with discrepancies for our best-fit depth (75 km, based on the optimum fit to predicted arrival times) ranging up to 2 s. This likely represents the three-dimensional nature of the near-source velocity structure, which is not well modelled, and is not accounted for in the one-dimensional velocity model used in predicting phase arrival times. This effect is rarely a significant problem with shallow earthquakes, as the velocity structure along the depth-phase raypath for stations on difference sides of the focal sphere is little different, but at the extreme depth of periosteum earthquake, depth phase bounce-points may be separated by tens of kilometers at the surface, which, in the case of this earthquake, can mean the difference between a depth phases passing through the basement-cored Wind River mountains, or through the sedimentary Wind River basement, with different velocity structures, and different elevations.
↧