Deep learning methods for modeling infrasound transmission loss in the middle atmosphere

2024-10-04

INTER-NOISE and NOISE-CON Congress and Conference Proceedings

AI Summary

To address the urgent need for real-time, high-accuracy modeling of infrasound transmission loss (TL) in the middle atmosphere for the International Monitoring System (IMS), this study overcomes the computational bottlenecks of conventional parabolic equation methods—prohibitive cost and infeasibility for global operational use (≤4000 km). We propose a novel multi-physics-coupled optimized convolutional neural network (CNN) model. For the first time, it enables end-to-end global TL prediction using joint inputs of temperature and 3D wind fields, substantially improving physical causality and generalization—particularly upwind and at low frequencies (0.1–4 Hz). Trained on an ultra-large-scale synthetic waveform dataset spanning regional and global scenarios, the model achieves a mean prediction error of only 20 dB with negligible inference latency per prediction. This work delivers the first IMS-compatible real-time TL modeling framework that simultaneously satisfies accuracy, computational efficiency, and physical consistency.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search

Abstract

Accurate modeling of infrasound transmission losses (TLs) is essential to assess the performance of the global International Monitoring System (IMS) infrasound network. Among existing propagation modeling tools, parabolic equation method (PE) enables TLs to be finely modeled, but its computational cost does not allow exploration of a large parameter space for operational monitoring applications. To reduce computation times, Brissaud et al. (2022) explored the potential of convolutional neural networks (CNNs) trained on a large set of regionally simulated wavefields (>1000 km from the source) to predict TLs with negligible computation times compared to PE simulations. However, this new method shows difficulties in upwind conditions, especially at low frequencies, and causal issues with winds at large distances from the source affecting ground TLs close to the source. In this study, we have developed an optimized CNN network designed to minimize prediction errors while predicting TLs from globally simulated combined temperature and wind fields spanning over propagation ranges of 4000 km. Our approach enhances the previously proposed one by implementing key optimizations that improve the overall architecture performances. The implemented model predicts TLs with an average error of 20 dB in the whole frequency band (0.1-4 Hz) and explored realistic atmospheric scenarios.

Problem

• Model infrasound transmission loss accurately
• Reduce computational cost of parabolic equation method
• Improve prediction in unfavorable wind conditions

Innovation

• Optimized convolutional network for global infrasound modeling
• Predicts transmission losses from combined temperature and wind fields
• Reduces average error to 8.6 dB across 0.1-3.2 Hz