Workshop Program
Register now
Workshop Program

Insight of the glacier dynamics in the region of Qaanaaq, Greenland, using infrasound data
by Dr. Sentia G. Oger, Dr. Patrick Hupe


Polar ice caps represent the largest reservoir of freshwater on Earth. In the context of climate change, we observe an increase of ice discharge in the Ocean. In Greenland, fractures (i.e. calving) and submarine melting of marine-terminating (tidewater) glaciers account for more than 80% of the Greenland ice loss. However, because such events are not comprehensively inventoried, and consequently the mechanisms at play are still unresolved, it is not possible to predict the future behaviour of these glaciers and its associated sea level rise. The dynamics of glaciers can generate acoustic waves that propagate through the atmosphere as infrasound.  Consequently, infrasound records could help in detecting ice discharge and documenting the associated mechanims, on top of existing deployed technologies. In the framework of the Comprehensive Nuclear-Test-Ban Treaty aimed to detect nuclear test explosions, an infrasound station was set up in Qaanaaq in Northwest Greenland, and certified in late 2003. We analyzed the continuous 17-years infrasound measurement recorded at Qaanaaq alongside atmospheric temperatures and winds simulated by the European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis ERA5, sea ice concentrations extracted from National Ocean and Atmospheric Administration (NOAA) satellite imagery products, as well as discharge outputs simulated by the regional climate model MAR. After a climatology study of the data, we used machine learning algorithms to better learn from the data. A first preliminary analysis of the infrasound data shows the existence of five source regions, distinguishable by the seasonality of the detections, highlighting the different processes at play for each of these regions. The seasonal comparison of the back-azimuth distribution gives hints on the potential sources, i.e. land glacier, tidewater glacier or sea ice. These results are confirmed when running the interpretable machine learning method random forest, fed with all the aforcementionned climate data. Once robustly identified, the next step will be to have a deeper insight into specific source events. This study highlights the potential of studying infrasound data for expanding our knowledge of the dynamics of glaciers.


Open in a new tab

Topic : Theme 1: Cryosphere monitoring.
Reference : T1-E16

Back to the list of submissions
Previous submission · Next submission


Comments are only accessible to participants. If you are a participant, please log in to see them.