The adequacy of routinely available data for identification of all earthquakes of mb ≧ 4.5 (Evernden, 1975) occurring throughout the world and listed in ISC bulletins for January 1, to June 10, 1972 is investigated. When using an identification criterion based on (1) depth ≧ 50 km when using P data alone, ≧ 30 when using pP — P or S — P data and/or (2) at least one of three long-period phases 8 or more times that of the largest value seen for an explosion, available data successfully identify all except two of the earthquakes studied (other than mixed events), these two being of magnitude 4.5. Only one of these two appears “anomalous” on the basis of available data. “Anomalous” earthquakes in nearly all cases must be the result of inadequate data or analysis and not similarity of spectral source function of these earthquakes and explosions.
Data on amplitudes of long-period S versus mb and depth are given, as well as further analysis of the mixed event problem. It is pointed out that the capability for detection of long-period waves here shown can be increased by half a magnitude by proper processing of presently available data.