Of the waves which propagate in the atmosphere at acoustic velocity in the period range from 10 to 100 s, one type has been classified by triangulation as arising principally from mountainous regions. These signals were first described as ' northwesters ' or ' 310 ers ' by the NBS Geoacoustics Group under R. K. Cook at Washington, D.C., from the predominant direction of arrival. Subsequent operation of an observatory at Boulder, Colorado by Vernon Goerke gave a source region by triangulation in the Pacific Northwest, primarily in Montana and Alberta. Installations of observatories at College, Alaska (Wilson) and Pullman, Washington-Moscow, Idaho (Craine and Thomas) enlarged the data base available, and triangulation showed the principal source areas to be along the coast of British Columbia and in the inland Rocky Mountains of the British Columbia-Alberta border. This paper discusses the presently known characteristics of this class of infrasonic waves, locates the triangulation areas, reviews selected events, and suggests that certain of these waves are produced as aerodynamic sound. The paper shows a correlation between the 500 mb jet stream velocity and direction in these mountainous regions, and the detection of these atmospheric pressure waves. Infrasonic waves-10to 100-s periodsThere are a number of atmospheric wave structures travelling at acoustic velocities that appear in the 10-to 100-s period range. An observatory, through four pressure sensitive transducers, provides basic information on the amplitude variations of pressure with time, from which the waveform, period, and time sequence of the event at an individual microphone can be obtained. By overlaying the charts and visually correlating the waveforms from several transducers, a signal can be identified in the noise, and arrival time differences measured. The azimuth of the direction of arrival, and the horizontal velocity across the transducer array, can then be calculated, assuming a plane wave. Detection of signals is often difficult, since local wind noise may obscure the desired signal, and changes in waveform caused by frequency dispersion of the signal or the effects of other signals arriving from different directions may make the waveform vary at each transducer. We define an ' event ' as a series of signals that appear to be related. 201 0 pr) FIG. 2. Correlation of mountain-associated waves at Pullman, Washington, at 1430 UT on 1968 November 28. Mean azimuth 329O, trace velocity 426 m s-l; approximate peak-to-peak amplitude 6 3 dyn cm-*; average period 50 s. facing page 202 Correlation of winds and geographic features 5 -203 15 10

Previous studies have identified a class of infrasonic waves characterized by periods ranging from 10 to 100 seconds, horizontal trace velocities across the detecting array at acoustic velocities or greater, and zero to peak amplitudes from 0.5 to 7 dyne cmm2. These signals triangulate principally in mountainous regions and have thus been termed mountain-associated waves. In this paper, the effects of propagation conditions on the observed characteristics are examined using a ray-tracing technique implemented on a hybrid computer. It is shown that the observed seasonal variation in occurrence of these waves follows from the conditions along the propagation path-primarily winds-and therefore may not be indicative of variations in the actual generating mechanism.

No abstract

Commonly occurring infrasonic waves recorded during the winter months are believed to originate by means of aerodynamic source mechanisms. Acoustic power spectra of these commonly observed signals are compared to the power spectra one would observe from theoretical source mechanisms. From this comparison, the theoretical source mechanism believed responsible for production of the observed signals is thought to be isotropic turbulence in the lee of mountain peaks. Experimentally recorded signals and their power spectra are shown. Source regions of three signals are identified.

No abstract