Computer modelling aids understanding of coastal risks

Scientists at RENCI and the University of North Carolina (UNC) at Chapel Hill, USA, are using the latest modelling techniques and high-performance computing power to understand how expected increases in sea level over the next 100 years could affect local coastal communities, wildlife and the coastline itself. Tom Shay, a UNC-Chapel Hill marine scientist, is using the ADCIRC coastal storm surge modelling software and the SWAN (Simulating WAves Nearshore) wave modelling software to create new ‘What If?’ models, showing how a major storm could affect the North Carolina coast if sea levels were higher.

The research complements work done at RENCI, which uses ADCIRC and SWAN to compute thousands of different possible storm scenarios.  The US Federal Emergency Management Agency (FEMA) uses the RENCI model output to develop new Flood Insurance Rate Maps (FIRMs) for coastal North Carolina as part of the North Carolina Floodplain Mapping Program. Shay’s work focuses on one storm: Isabel, which made landfall in North Carolina in 2003 as a Category 2 hurricane. Using data on coastal topography and bathymetry compiled for the floodplain mapping modelling effort, he models the storm at current sea levels and at sea levels ranging from 0.5 to 2 metres higher.

Although his simulations show more pronounced storm surge and flooding during the hypothetical Hurricane Isabels, Shay stressed that the results depict a possible future, not a certainty. ‘There are a lot of unknowns and uncertainties,’ he said. ‘A warming ocean will cause some sea level rise just because of thermal expansion, but we don’t know how fast the ocean will warm or how fast the Greenland ice will melt or whether the West Antarctic ice sheet will melt. All those things have a range of possible outcomes and we want to understand as many of the possible outcomes as we can.’

RENCI senior scientist Brian Blanton develops simulations using the ADCIRC and SWAN models, but focuses on understanding the risks coastal communities will face from hurricanes and extratropical storms if sea levels rise. The work, funded by the North Carolina’s Floodplain Mapping Programme, uses the same datasets as the floodplain modelling project and data from Applied Research Associates on probable hurricanes that could head toward North Carolina over the next 100 years. Blanton then uses that data to generate probable hurricane and severe storm events under different conditions that could be affected by climate change, such as higher sea levels, increased storm intensities and changes in the frequencies of storms.

The work will harness the power of a RENCI supercomputer to run about 2,000 individual storm simulations in six different climate scenarios, according to Blanton. Each simulation set will be crunched into statistical analyses that will help to clarify how coastal risks change under each climate scenario. The end product, he said, will be a scientific evaluation of the risk of living on the coast—whether economic risk, risk to infrastructure or risk to lives—under changed conditions, which planners, emergency managers, scientists and policymakers will then be able to compare to their current understanding of coastal risks.

Twitter icon
Google icon icon
Digg icon
LinkedIn icon
Reddit icon
e-mail icon
Analysis and opinion

Robert Roe looks at research from the University of Alaska that is using HPC to change the way we look at the movement of ice sheets


Robert Roe talks to cooling experts to find out what innovation lies ahead for HPC users

Analysis and opinion