CalvingSEIS: Glacier dynamic ice loss quantified through seismic eyes

About 40% of global glaciers and ice caps, excluding ice sheets of Greenland and Antarctica, loose mass through iceberg calving. This dynamic ice loss is a sensitive component of the mass transfer from glaciers to oceans through a number of feedback processes with climate. Current models are currently not equipped to realistically predict dynamic ice loss, mainly because long-term, high temporal resolution and continuous calving records are inexistent. CalvingSEIS aims to produce continuous calving records using combined passive seismic/acoustic strategies; the only technique able to capture rapid calving events, continuously, and back through time over decades. CalvingSEIS focuses on the glaciers in Kongsfjord, Svalbard as the research village of Ny Ålesund houses a passive seismic instrument since 1994 and is only 15 km from one of the fastest flowing and most heavily studied glaciers in Svalbard, Kronebreen. Archive time-lapse imagery from 2009/10 has been re-processed to calibrate a seismic calving detector from the permanent seismometers in Ny Ålesund and Longyearbyen. A 5 year satellite record of glacier frontal ablation (calving ice plus submarine and subaerial melting of the calving front) at monthly temporal resolution is used to calibrate seismic detections of calving events into calving volume loss. This produced a 15 year calving flux record for Kronebreen with a weekly resolution at best. To our knowledge, this is the first ever long term continuous calving flux estimate from any glacier globally, and the first ice loss record derived from passive seismic records. Over 14 days in August/September 2016, a calibration field experiment occurred at the front of Kronebreen. With 14 people from 6 nations and 9 nationalities, we measured processes and properties of individual calving events. Over 1500 calving-like signals were recorded on passive seismic and underwater acoustic instrumentation during this two week period. Time-Lapse imagery provide movies of the calving front and repeat terrestrial laser scanning measured calving event volumes. Two radars operated synergistically to capture velocities every 15 minutes. All together, these records provide an amazing mechanical history of the calving front. Results indicate the possibility to predict calving events up to days before the actual ice calving using such instrumentation. Furthermore, using the calving volumes of 100 calving events, we are able to define a relationship between the logarithmic of calving volume and seismic signal characteristics, enabling a direct calibration and transfer function from calving event detections to ice volumes. Using this relationship, we calculate the ice flux of Kronebreen based upon our direct calibration and find that this estimate is about 1/3 of our calibrated estimate from satellite image reconstruction of frontal ablation. This implies that submarine melt may account for up to 2/3 of the ice loss from the calving front, and further that this may be the first order control on calving. In summary, passive seismology is a remarkable resource in glaciology for observing the temporal variation of very rapid processes, such as calving. The potential to use regional passive seismic stations to quantify glacier frontal ablation are applicable in many other glaciated regions. The dynamic ice loss timeseries produced here are globally unique providing an encyclopedia and data bank for modellers, helping contribute to defining Kronebreen as a benchmark glacier for calibrating and validating models of ice dynamics and calving. Through this project, passive seismology has also been used to detect and observe glacier dynamic instabilities such as surges in Svalbard. Combined with satellite data and thermo-mechanical modelling, we are able to confidently show that the 14000 seismic events during the Nathorstbreen collapse resulted from ice tearing after destabilization of the frozen glacier tongue. Further pilot studies indicate the ability of passive seismics recorded within glaciers to record surface crevassing and basal processes as well as the ability to observe seasonal variations in active layer thickness of permafrost, both interesting topics for future research. Bioacoustics recorded in the marine environment has a similar potential. We have not only detected and calibrated calving from Kronebreen using these instruments, but localized and analyzed seal and whale behavior in relation to external factors such as sea ice and ocean currents. The use of bioacoustics for understanding subsea melting through bubble release remains difficult due to the presence of sea ice, fjord ice, icebergs and the calving glacier face. Nonetheless, combined approaches using satellite remote sensing, passive seismology and marine acoustics provides an immense potential to better understand the physical and biological interactions between tidewater glaciers, their fjord environments and the atmosphere.

-

About 40% of global glaciers and ice caps (excluding Greenland and Antarctica) loose mass through iceberg calving, while current models (e.g. in the IPCC) are currently not equipped to realistically predict dynamic ice loss, mainly because long-term continuous calving records are inexistent. Combined seismic/acoustic strategies are the only technique able to capture rapid calving events, continuously, and back through time over decades. CalvingSEIS aims to produce high temporal resolution, continuous calving records for the glaciers in Kongsfjord, Svalbard, and in particular for the Kronebreen glacier laboratory. Through innovative, multi-disciplinary monitoring techniques combining fields of seismology and bioacoustics, individual calving events will be detected and located autonomously and methods to quantify calving ice volumes directly from the seismic and acoustic signals will be developed. Scaling seismic/acoustic signals to calving ice volumes require calibration. CalvingSEIS will generate a detailed catalogue of calving events employing state-of-the-art terrestrial remote sensing techniques to measure calving ice volumes, velocities, and ice-ocean interactions. Experimental terrestrial remote sensing observations and investigations will further invoke process-based understanding at the transition zone between glacier and ocean. The underwater bio-acoustic sensors will collect not only glacier sounds, but record the entire fjord soundscape instigating studies between biotic, abiotic and anthropogenic components; e.g. marine animal interaction with glacier sounds from calving and melting. The main project goal to generate a dynamic ice loss timeseries will reveal fine scale processes and key climatic-dynamic feedbacks between glacier calving with climate history, topographic setting, terminus evolution and fjord conditions. This timeseries will form an unprecedented dataset for developing, calibrating and validating glacier dynamic models.