Assessing and optimizing the performance of infrasound networks to monitor volcanic eruptions

We propose a numerical modeling technique based on a frequency-dependent attenuation relation to assess, quantify and optimize the performance of any arbitrary infrasound network to monitor explosive sources such as volcanic eruptions. Simulations are further enhanced by including realistic sources and propagation effects. We apply our approach to both hemispheres by considering the Euro-Mediterranean and the Eastern Australian regions. In these regions, we use quasi-permanent infrasound signals from Mt. Etna recorded in Tunisia and from Mt. Yasur recorded in New Caledonia. These well-instrumented volcanoes offer a unique opportunity to validate our attenuation model. In particular, accurate comparisons between near and far-field recordings demonstrate the potential of the proposed methodology to remotely monitor volcanoes. A good agreement is found between modeled and observed results, especially when incorporating representative 10 m/s wind perturbations in the atmospheric specifications according to previous campaign measurements. To optimize the network layout in order to ensure the best monitoring of the volcanoes, we proceed through a grid search to find optimum locations of an additional array. We show that adding one array at an appropriate location in both regions under study could significantly improve detections half of the year. The application of the proposed methodology can provide in near real-time a realistic confidence level of volcanic eruption detections, useful to mitigate the risk of aircrafts encountering volcanic ash.