Observations of atmospheric nuclear tests carried out at the Nevada Test Site from 1951 to 1958 provided data for an empirical investigation of how infrasonic signals are propagated to distances of about 250 km. Those observations and the analysis documented in this report involved signal amplitudes and average velocities and included three classes of signals: stratospheric, thermospheric, and tropospheric/surface. The authors' analysis showed that stratospheric winds have a dominant effect upon stratospheric signal amplitudes. The report outlines a method for normalizing stratospheric signal amplitudes for the effects of upper atmospheric winds and presents equations for predicting or normalizing amplitude and average velocity for the three types of signals.