Noise Characteristics of the Alaska Regional Infrasound Network and Implications for General Infrasound Station Performance

The Alaska Regional Infrasound Network consists of 107 single infrasound sensor-equipped stations situated across a large swath of the state of Alaska with roughly 85 kilometer irregular spacing (see Fig. 1). These sites are a legacy of the USArray Transportable Array deployment to Alaska, and were adopted into the regional seismic network (AK) by the Alaska Earthquake Center and are being analyzed by the Wilson Alaska Technical Center, both of the University of Alaska Fairbanks. While deployments of infrasound sensors colocated with seismometers are becoming common, single-sensor regional infrasound networks are rare. Because these stations were sited without regard to acoustic performance and span a geographically diverse region, they provide a unique opportunity to evaluate the effects of site conditions on infrasound noise performance. Colocated with the infrasound at 70 of these sites is a Vaisala meteorological station, allowing for the examination of the effects of wind on noise. In this study, we use a year of infrasound and wind data from 2020 to gain insight into the acoustic environment and performance of the network. For each site, we calculated power spectral density curves every 4 hours and compiled an ambient acoustic model for the Alaska region for use as a data quality tool. Due to the diversity of geographic site conditions, this model may be applicable globally. In order to investigate long-term noise characteristics and resolve diurnal and seasonal effects at each station, we then used the power spectral densities to compute year-long time series of root mean squared amplitudes in several passbands. From these observations, we extracted models for estimating noise amplitudes as a function of wind speed and site land cover classification. Our results indicate that ambient noise levels for the network are highly dependent on land cover type and that wind is a major source of noise across a broad frequency band. We also observed that noise amplitudes are dependent on wind direction at several sites. We conclude that noise-wind relationships that consider land cover type are highly useful for infrasound station site selection and prediction of expected noise levels. This work was supported by the Nuclear Arms Control Technology (NACT) Program at Defense Threat Reduction Agency (DTRA). Cleared for Release.