The objective of this research is to make a systematic analysis of intermediate and deep earthquakes with large non-double-couple source components. The primary task is to determine from previously identified non-double-couple events whether they are "intrinsically" non-double couples, or whether instead they appear to be sums of subevents which individually appear to be double couples. Some events will be studied in a matched pair mode where two events from the same area differ markedly in their non-double-couple components in order to minimize path effect and other systematic differences. True non-double-couple events will be classified for location and other systematic characteristics. This work has direct application to the problem of the source of deep earthquakes.
