This report deals with the problems of variation of Pn amplitudes in the regional and near-regional distance ranges (200-2100 kilometers). The data used were recorded by Long Range Seismic Measurement vans of the VELA Seismological Center as a result of earthquakes throughout the United States and numberous nuclear and chemical explosions in the same region.
It is shown that the patterns of Pn amplitudes versus Δ in Western United States and Eastern United States are markedly different and that these differences are related to different velocity structures in the two regions. These differences extend to at least 150 kilometers depth.
Neither the Western United States nor Eastern United States amplitude patterns conform even approximately to that predicted or suggested by Gutenberg and Richter in the 1000-2000 kilometer range. The predicted zone of low-amplitude signals in that distance range does not exist and overestimation of magnitude by as much as 1.5 magnitude units is frequently done because of failure to properly understand the patterns of radiation.
By proper calibration of the Western United States by use of numerous events, it is now possible to make consistent estimates of magnitude at all distance ranges for most explosions and earthquakes. Obtaining consistent estimates of amplitude as a function of distance for a particular event requires a knowledge of the energy distribution between the several refracted phases used between 200 and 2500 kilometers distance. Data on hand show that the energy partition function is reasonably uniform throughout the regions investigated but that locally it may vary radically, resulting in a ten-fold change in relative excitation of two refracted phases.
Patterns of energy radiation and decay are probably approximately determinable from knowledge of velocity structure and vice versa. Both are best determinable from explosion data where origin time and focus are accurately known.