Using an oceanic acoustic noise model to evaluate simulated atmospheric states

Infrasound of oceanic origin, known as microbaroms, are globally and continuously detected by the infrasound stations of the International Monitoring System. They propagate over thousands of kilometers thanks to acoustic waveguides in the middle and upper atmosphere (~30-120 km). At these altitudes, Numerical Weather Prediction (NWP) models are biased, partly due to the lack of operationally assimilated observations that constrain model predictions (especially for winds). We present a processing chain that simulates microbarom arrivals at infrasound stations by coupling a microbarom source model and infrasound propagation modelling. These arrivals are modelled using atmospheric specifications from different NWP models and compared to the microbarom observations using a multidirectional metric. The objective of this processing chain is twofold: evaluating the relative performances of NWP models in support of operational infrasound monitoring activities and demonstrating the benefit of assimilating unconventional observations such as microbaroms in NWP models. To this end, we apply the processing chain to several infrasound stations and highlight NWP performance assessments during a sudden stratospheric warming and other dynamical events of the middle and upper atmosphere.