In this paper, an empirical method is developed for determining magnitude of seismic events from the amplitude of Lg phases recorded at local to regional distances from the events. The data used were obtained from Long Range Seismic Measurements (LRSM) reports for 78 seismic events associated with explosions conducted the Atomic Energy Commission.
Preliminary investigation showed that magnitudes (M) calculated from Lg recordings had less scatter than body-wave magnitudes (m) reported for the same events and stations. A sixth-degree polynomial was found to give the best fit to log amplitude/period (A/T) values plotted as a function of epicentral distance. Amplitude-distance correction factors (Q) are presented in a chart.
Magnitudes from Lg were compared to the average unified magnitudes from LRSM shot reports. For 60 of 68 events, the standard deviation for M was less than that for m. The M values for all of the events were normalized to magnitude 5.0, and average residuals were found for 69 reporting stations. A plot of the residuals on a map indicates that they represent station corrections rather than azimuthal variations. Apparent station corrections (s) are presented for 23 LRSM stations.