Swiss scientist builds new model to simulate tsunami generation from iceberg calving

Johan Gaume, an EPFL expert in avalanches and geomechanics, has turned his attention to ice. His goal is to better understand the correlation between the size of an iceberg and the amplitude of the tsunami that results from its calving. Gaume, along with a team of scientists from other research institutes, has just unveiled a new method for modeling these events. 

These scientists are the first to simulate both glacier fracture and wave formation phenomena when the iceberg falls into the water. “Our goal was to model the explicit interaction between water and ice – but that has a substantial cost in terms of computing time. We, therefore, decided to use a continuum model, which is very powerful numerically and which gives results that are both conclusive and consistent with much of the experimental data,” says Gaume, who heads EPFL’s Snow Avalanche Simulation Laboratory (SLAB) and is the study’s corresponding author. The other institutes involved in the study are the University of Pennsylvania, the University of Zurich, the University of Nottingham, and Switzerland’s WSL Institute for Snow and Avalanche Research. Eqip Sermia glacier and its icebergs. © iStock

Improving calving laws
The scientists’ method can also provide insight into the specific mechanisms involved in glacial rupture. “Researchers can use the results of our simulations to refine the calving laws incorporated into their large-scale models for predicting sea-level rises while providing detailed information about the size of icebergs, which represent a sizeable amount of mass loss,” says Gaume.

Calving occurs when chunks of ice on the edge of a glacier break off and fall into the sea. The mechanisms behind the rupture generally depend on how high the water is. If the water level is low, the iceberg breaks off from the top of the glacier. If the water level is high, the iceberg is longer and breaks off from the bottom, before eventually floating to the surface owing to buoyancy. These different mechanisms create icebergs of different sizes – and therefore waves of different amplitudes. “Another event that can trigger a tsunami is when an iceberg’s center of gravity changes, causing the iceberg itself to rotate,” says Gaume. “We were able to simulate all these processes.”

In Greenland, the scientists placed a series of sensors at Eqip Sermia, a 3-km-wide outlet glacier of the Greenland ice sheet that ends in a fjord with a 200 m ice cliff. Back in 2014, an iceberg measuring some 1 million m3 (the equivalent of 300 Olympic-sized swimming pools) broke off the front of the glacier and produced a 50 m-high tsunami; the wave was still 3 m high when it reached the first populated shoreline some 4 km away. The scientists tested their modeling method on large-scale field datasets from Eqip Sermia as well as with empirical data on tsunami waves obtained in a laboratory basin at the Deltares Institute in the Netherlands.

Projects in the pipeline
Glacier melting has become a major focus area of research today as a result of global warming. One of the University of Zurich scientists involved in the study kicked off a new research project this year with funding from the Swiss National Science Foundation. This project will investigate the dynamics of Greenland's fastest-moving glacier, Jakobshavn Isbrae, by combining data from individual field experiments in Greenland with the results of simulations run using the SLAB model. “Our method will also be used to model chains of complex processes triggered by gravitational mass movements, such as the interaction between a rock avalanche and a mountain lake,” says Gaume.

Intel's chip sales plunge 20 percent in Q1

Intel has reported a huge drop in data center chip sales and a steep decline in gross profit margin, a sign it’s losing market share to rivals and customers who are designing their components. Intel shares were down 3% in after-hours trading after the results.

The company’s Data Center Group (DCG) has generated first-quarter 2021 sales of $5.6 billion, down 20% from a year earlier and below Wall Street estimates. This is one of Intel’s most profitable businesses, so the lower revenue hurt the overall margins.

In the fourth quarter of 2020, Intel's DCG revenue amounted to approximately $6.09 billion, a decrease of 15% from the figure reported for the same quarter of 2019.

Intel said sales of chips to cloud computing service providers fell 29% from the same period a year earlier. That big drop, according to Intel, was caused by “digestion” customers pausing orders while they work through unused stockpiles of chips.

Demand for cloud computing services from data centers has surged during the pandemic as many businesses shifted to working from home.

“This is a pivotal year for Intel. We are setting our strategic foundation and investing to accelerate our trajectory and capitalize on the explosive growth in semiconductors that power our increasingly digital world,” said Pat Gelsinger, Intel CEO.

SUNY Stony Brook's Reed forecasts attribution of the human influence on Hurricane Florence

A study led by Kevin Reed, Ph.D., Assistant Professor in the School of Marine and Atmospheric Sciences (SoMAS) at Stony Brook University, and published in Science Advances, found that Hurricane Florence produced more extreme rainfall and was spatially larger due to human-induced climate change.

Previous research has suggested that human influences such as emission of greenhouse gasses that alter climate do affect precipitation in extreme storms. The research in this study, however, is a first to use a “forecast attribution” framework that enables scientists to investigate the effect of climate change on individual storm events days in advance. Kevin Reed, PhD{module INSIDE STORY}

In 2018, prior to the landfall of Hurricane Florence, Reed and colleagues made predictions based on simulations of the storm given climate change models. They predicted Hurricane Florence would be slightly more intense for a longer portion of the forecast period, rainfall amounts over the Carolinas would be increased by 50 percent due to climate change and warmer water temperatures, and the hurricane would be approximately 80 kilometers larger due to the effect of climate change on the large-scale environment around the storm.

“With our ability for additional ‘hindsight’ numerical modeling of the storm around climate change factors, we found predictions about increases in storm size and increased storm rainfall in certain areas to be accurate, even if the numbers and proportions are not exact,” explains Reed. “More importantly, this post-storm modeling around climate change illustrates that the impact of climate change on storms is here now and is not something only projected for our future.” 

Reed explains the future tools of storm modeling including the approach by his team in this brief video.  

{media id=235,layout=solo}

He said that while the post-storm analysis did show that the storm was slightly more intense during the forecast period due to climate change – as they predicted – as measured by minimum surface pressure and near-surface winds, the finding remains the most uncertain from the hindsight model.

One key finding of the post-storm model showed that Hurricane Florence was about nine kilometers larger in mean maximum diameter due to climate change. Additionally, rainfall amounts over large ranges were significantly increased. Mean total overland rainfall amounts associated with the forecasted storm’s core were increased by 4.9 ± 4.6% with local maximum amounts experiencing increases of 3.8 ± 5.7% due to climate change.

Reed emphasizes that by attributing climate change effects to individual storms, as his team did with Hurricane Florence, scientists are better able to communicate the direct impacts of climate change on extreme weather to the public.