Direct numerical simulations of long-range infrasound propagation: Implications for source spectra estimation.

The evolution of observed dominant frequencies from a high-intensity infrasonic pulse with receiver range and stratospheric temperature is investigated using direct numerical simulations of the two-dimensional unsteady compressible Navier-Stokes equations. There is a high level of uncertainty in estimating source dominant frequencies based on received signals at sparse points on the ground. Nonlinear propagation effects in the ground-level thermospheric arrivals are found to significantly alter dominant frequency measurements compared to stratospheric arrivals with smaller amplitude sources. With a larger amplitude source, variations in observations are minimized as a result of nonlinear effects being ubiquitous across all atmospheric components of received signals but have a greater offset to the source dominant frequency. An approach to determine the source dominant frequency and minimize atmospheric variability is presented by calculating a source-to-receiver spectral transfer function averaged across the atmospheric states. This method reduces atmospheric variability in source frequency estimates within the pseudo-linear propagation regime and the average error to the known source frequency with a large amplitude source. The reduction of errors in source frequency estimates demonstrates the feasibility of using remote infrasound measurements as an indicator of source frequency and, in turn, the explosive yield of clandestine nuclear weapon test explosions.