An Analysis of Unmanned Aerial Vehicle-Based Acoustic Atmospheric Tomography

The engine of an unmanned aerial vehicle generates acoustic energy comprising multiple prominent narrowband tones superimposed onto a broadband background. This signature may be synchronously observed both onboard the aircraft and by a set of microphones on the ground. Comparison of the projected and observed signals allows computation of atmospheric sound speed values for a set of intersecting rays. Tomography may then be used to estimate spatially varying temperature and wind velocity profiles of the intervening medium. Errors in these atmospheric reconstructions are a function of several factors, which include: refraction, multi-path, atmospheric turbulence, position and velocity errors, instrumental errors in the measurement of the signals, and statistical uncertainty in processing techniques used to perform the signature comparisons. Vertical cross-sections of weakly-sheared daytime convective atmospheric boundary layers, created using large eddy simulation code, are combined with error models of the geometry, propagation and instruments to determine levels of inaccuracy that can be tolerated for faithful reconstruction of target atmospheres