Digging deeper in dynamics of ice

Icebergs calved from glaciers in Greenland. Picture: Yongmei Gong

Digging deeper in dynamics of ice

Maria Virkkula

Glacial ice covers approximately 10 percent of the global land surface, and melt water flows under some of the condensed ice. During the past millennium glaciers around the world have rapidly retreated and shrunk, and polar region ice mass flow has accelerated. 

Yongmei Gong (in the picture), from the University of Helsinki, explores how ice sheets in Antarctica will evolve in next 100 to 200 years under climate change and what has triggered the mechanisms of fast ice flowing events, surges, in an Arctic ice cap. This is investigated in her fresh doctoral thesis on modeling of ice flows in ice shelves and ice caps.


Numerical simulations provide a way to study the physical processes of ice. A major part of Gong's research is done with Elmer/Ice, a software for ice sheet, glacier and ice flow modeling, mainly developed at CSC. Additionally, she has used an ice sheet model BISICLES developed by the researchers in Britain and the U.S. as well as a discrete particle model which is also developed in CSC.

Numerical models allow an accurate, high-resolution view of processes like glacier surges, ice streams and grounding-line migration.

– All the known physical processes can be described by mathematical formulations to certain extent. My research goal is to use the models to better understand and explain certain physical processes and natural phenomenon in the real world. The specific aim is to use the state-of-the-art three-dimensional model Elmer/Ice to investigate the acceleration of the ice flow in Basin 3, Austfonna Ice-cap, Gong says.

Everything started from a dream to travel to Antarctica

Austfonna is one of largest ice caps in Europe located on Nordaustlandet in the Svalbard archipelago, Norway. Basin 3 is one of its south-eastern glaciers. It is a fast-flowing surging glacier, where seasonal accelerations interrupt the year-round slower long-term flow.

The glacier in Basin 3 is now flowing much faster than before. Also, step-wise multiannual acceleration, associated with seasonal summer speed-up events, occur.

– With the help of two ice flow models and a discrete particle model, I have tried to explain the mechanisms that have affected the acceleration of the normal glacier movements since 2011.

The discrete particle model divides glaciers to smaller particles, and provides the possible location of the crevasses that can potentially cut thought the full length of ice and deliver surface melt water down to the bed.

Gong's results indicate that the summer speed-up events and the acceleration in the basin could be explained by either a direct enhancement to the ice flow through basal lubrication or the outburst of the water accumulated in a topographically over-deepened area later during the acceleration.

– I did similar research on an ice shelf in Antarctica for my master's thesis. I had no idea what numerical modeling is, until I moved to England and started to work with ice dynamic models. I thought if I did something about Antarctica, I would have a chance to go there, Gong admits.

– I have not given up my hope even though, later, I realized that doing ice flow modeling basically means sitting in front of your computer and writing codes all the time.

A struggle that led to a new approach

When Gong started her PhD, the research questions were not clearly defined. She spent nearly two years getting advice from senior researchers and trying all kinds of numerical experiments. Through her master's theses she entered deeper into the world of glaciology and numerical modeling. And now, she's only getting started.

– My PhD is a natural continuation and resulted from some practical factors, e.g. funding, the topic was interesting and the supervisors were well known etc. At the beginning, I had the numerical model and the data, but I didn't have a good idea. Finally, with the help of my supervisors and colleagues, I came up with this idea of using both the ice dynamic model and the discrete element model. This approach has not been done before. But I guess this kind of struggle is normal for all scientific research. You need to keep experimenting and brainstorming with other researchers.

Yongmei Gong doing field work in Greenland.

"Large-scale ice dynamic models need to be developed"

On top of Austfonna Ice cap, Gong investigates one of the biggest drainage systems in East Antarctica, Lambert Glacier – Amery Ice Shelf drainage system (LG-AIS). Amery Ice Shelf is going through a steady cycle of producing icebergs into the ocean. Due to the climate change the future of the drainage system is, however, unknown.

– The study suggests that even though large amount of melting have been predicted at the base of Amery Ice shelf, the LG-AIS drainage system is going to be rather stable in this century or the next century. The sea level contribution will be small, Gong concludes.

Nevertheless, by 2100, most large coastal cities are predicted to face sea levels that are more than a meter higher than currently, and the total effect of melting ice mass on sea level rise is widely debated.

– The effect of ice sheet and glacier melting on sea level rise has been talked about and – studied since long time ago. I suggest to read the comprehensive prediction in the IPCC 5th assessment report. You will notice that the range of the future sea level contribution from ice sheets and glaciers is wide, Gong says.

One reason being that there are still large uncertainties in future projections due to ice dynamic response to climatic forcings.

– Therefore, large-scale ice dynamic models need to be developed to better understand certain key physical processes regarding ice dynamic change. These key problems include but are not limited to: grounding migration, fast ice motion related to basal processes and linking climatic signal to abrupt changes of ice dynamics. At the same time, we also have a large amount of observational data. It is important to use numerical models that can handle large data sets to sufficiently investigate the problem in more detail.

– This also gives me the motivation to do ice dynamic modeling, Gong says.

Gong's PhD thesis solves one piece of a puzzle. Next, she is planning to find a research positions to explore the interaction between the glacial system and other systems such as ice-ocean or ice-atmosphere interaction. What lies ahead when technologies and methods evolve, remains to be seen.

– I think I am still too junior to make any grand forecast for the future of glaciology. Since the computational ability of super clusters has become so strong, and we have developed all kinds of numerical models describing different components in the earth system, incorporating different models together (coupling) and investigating the long-term interaction, for instance, between the glacial system and the atmosphere and the ocean could be an interesting direction to go. And people have already started to do so.


You might be also interested in:


comments powered by Disqus