Reconstruction de l'épaisseur de la glace de mer par inversion de forme d’onde d’icequakes et tomographie

Sea ice plays a crucial role in the Earth's climate system, serving both as a driver of and a sensitive indicator for environmental change—particularly in polar and sub-polar regions. This study seeks to improve our understanding of sea ice structure and its temporal variability in response to environmental forces by reconstructing sea ice thickness through seismic waveform inversion and tomography, using both passive and active seismic datasets. The broader goal is to contribute to a better understanding of climate changes.We use forward modeling of seismic wave propagation in sea ice to generate synthetic waveforms, employing the Spectral Element Method (SEM). These synthetic waveforms are then used in an inversion framework that compares them to observed waveforms recorded by seismic receivers deployed on the ice. By minimizing the misfit between synthetic and observed data through a stochastic approach combined with grid search, we estimate ice thickness along source-receiver paths and determine icequake locations. This inversion strategy is applied to both passive and active seismic source data. Once ice thickness estimates are obtained, they are used as input for a second inversion step: seismic tomography. This process yields spatially resolved maps of sea ice thickness.The approach is validated through synthetic tests and applied to several real datasets. The primary dataset, acquired on Lake Vallunden, Svalbard (Norway) between 28 February and 26 March 2019, includes continuous passive recordings as well as two active-source experiments, from 247 sensors. Daily icequakes extracted from passive data using machine learning algorithms form the basis for our inversion. Additional datasets from small-scale active-source experiments conducted in Québec, Canada, in March 2024 and February 2025 provide further validation of our approach. In all cases, seismic tomography results are consistent with field observations, including in-situ measurements and temperature records.The daily ice thickness maps derived from the Svalbard data reveal a clear spatial pattern, with thinner ice at the center of the lake and thicker ice toward the edges. They also capture daily changes in ice growth and thinning over time, showing meaningful correlations with observations and temperature fluctuations. These results demonstrate the robustness and sensitivity of our method, which combines effective numerical modeling with advanced inversion strategies. By capturing the daily evolution of sea ice structure, this work offers valuable insights into the structure and evolution of sea ice thickness, and introduces a novel tool for observing and understanding climate-driven changes in cryospheric environments.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.