Computational modelling of nonlinear infrasound propagation in the atmosphere

Abstract

In the field of numerical modelling for atmospheric infrasound propagation, linear acoustics prevails as a commonly used approximation, overshadowing the limited exploration of nonlinear effects in comparison to their linear counterparts. Moreover, there is a scarcity of numerical models specifically designed for studying nonlinear effects, with existing endeavours often relying on computationally intensive solutions for fluid dynamics equations. A key question emerging from this landscape is the practical impact on determining source properties from remote infrasound observations. This thesis addresses this gap by conducting investigations into the influence of nonlinear propagation on the remote sensing of source properties. Additionally, it introduces novel low-fidelity models to enhance the understanding of nonlinear effects in realistic atmospheres, serving as a benchmark for low-fidelity numerical solutions. In-depth analysis of ground-based infrasound signals generated through Direct Numerical Simulations are presented to assess uncertainties in source dominant frequency estimates, considering various approximations related to numerical propagation modelling and source determination methods. A novel approach is introduced to determine the source dominant frequency and mitigate atmospheric variability by calculating a source-to-receiver spectral transfer function averaged across atmospheric states. This method proves effective in reducing atmospheric variability and minimising errors in source frequency estimates. These analyses demonstrate the feasibility of obtaining source frequency estimates from remotely detected signals. Neglecting nonlinear effects in transfer function calculations also results in substantial errors in source dominant frequency estimates, demonstrating the impact of linear acoustics approximations in remote sensing applications. A model for nonlinear infrasound propagation through weakly varying, moving atmospheres is also devised. This analysis is shown to provide insights into the anticipated differences between nonlinear propagation effects in stationary and moving media. Direct Numerical Simulations of infrasound propagation through stratospheric winds and analogous no-wind atmospheres are performed, and wind-directional dependencies for nonlinear propagation effects are shown to align with predictions from the simplified model. Approximations of source dominant frequencies are made, revealing significant errors induced by common linear, low-fidelity acoustic modelling approximations. Nonlinear effects on remote sensing demonstrated in this thesis have profound implications for the remote detection and characterisation of clandestine nuclear weapon test explosions. To this end, efforts towards the development of low fidelity nonlinear propagation models present scope for further research into a wide range of atmospheric and source effects on remote characterisation with efficient calculation methods.

More information

Identifiers

ORCID for Liam James Tope: orcid.org/0000-0003-3426-009X

ORCID for Jae Wook Kim: orcid.org/0000-0003-0476-2574

ORCID for Alan Mcalpine: orcid.org/0000-0003-4189-2167

Catalogue record

Export record