Finite difference synthesis of infrasound propagation through a windy, viscous atmosphere: Application to a bolide explosion detected by seismic networks

A finite-difference time-domain (FDTD) algorithm has been developed to model linear infrasound propagation through a windy, viscous medium. The algorithm has been used to model signals from a large bolide that burst above a dense seismic network in the US Pacific Northwest on 19 February, 2008. We compare synthetics that have been computed using a G2S-ECMWF atmospheric model to signals recorded at the seismic networks located along an azimuth of 210° from the source; the results show that the timing and the range extent of the direct, stratospherically ducted and thermospherically ducted acoustic branches are accurately predicted. However, estimates of absorption obtained from standard attenuation models (Sutherland-Bass) predict much greater absorption for thermospheric returns at frequencies greater than 0.1 Hz than is observed. We conclude that either the standard absorption model for the thermospheric is incorrect, or that thermospheric returns undergo non-linear propagation at very high altitude. In the former case, a better understanding of atmospheric absorption at high altitudes is required; in the latter, a fully nonlinear numerical method is needed to test our hypothesis that higher frequency arrivals from the thermosphere result from nonlinear propagation at thermospheric altitudes.