The use of a 3-D Model to Improve Regional Event Locations

The appropriate role for 3-D models of the crust and upper mantle nuclear monitoring remains contentious. We argue that continent- and even global-scale 3-D models are emerging with sufficient reliability and resolution to warrant a central role in locating events using regional phase data alone. Sparse network locations with regional data are crucial in nuclear tests monitoring. We use a current model of the crust and upper mantle beneath Eurasia that we constructed by using surface-wave dispersion data. We present the results of a validation test in which, using regional phase data alone, we relocate more than 150 earthquakes and nuclear explosions from 15 reference event clusters in the Middle East and Central Asia. The location with the 3-D model is compared with the Ground Truth (in most cases GT5 or better) either known exactly or validated by cluster analysis. We find that the 3-D model improves locations relative to the 1-D model ak135 in 70-85% of all cases with the details depending on the number of reported phases used for relocation. Typically, the 3-D model reduces the location errors to about half the values attained with the 1-D model. This test indicates that the location of regional events is significantly improved by using a 3-D model for the region, even if the model is constructed on a much larger scale.