Three-dimensional topographic effects on infrasound propagation across Ascension Island

Abstract

SUMMARY Narrowband harmonic infrasound signals within the 1–8 Hz passband, generated by wind turbines on Ascension Island, have been recorded at four microbarometers located at distances of between 1.8 and 4.6 km from the source along different azimuths. Across one month of recordings in October 2010, amplitude ratios between the four recordings show temporal stability but deviate from the ratios expected for propagation across a flat plane. Using a recently developed 3-D parabolic equation method, that can incorporate realistic topography as a lower boundary, it is shown that these time-independent amplitude ratio deviations can be, in part, explained by acoustic interactions with topography that has scale lengths on the order of a few hundreds of metres. These interactions comprise both 2-D barrier effects that reduce sound levels behind high topography, and 3-D diffractive effects that increase sound levels behind topographic obstacles. For the Ascension Island case study, amplitudes along two of the four paths can be successfully modelled using a 2-D model, indicating that barrier effects dominate for these path geometries. Amplitude ratios along a third path, and the frequency-dependence of these ratios, are better simulated using a 3-D model that captures the out-of-plane diffractive effects around a prominent hill. The fourth path is poorly modelled using the 3-D model, which overpredicts acoustic amplitudes in this case. We hypothesize that this mismatch is likely to be due to a simplified description of the wind turbine source term. This study provides further observational confirmation that topographic interactions need to be considered when interpreting locally propagating infrasound, and shows that for harmonic narrowband sources a parabolic equation solver incorporating realistic boundary conditions provides an efficient method for simulating topographic interactions.