New model harnesses supercomputing power for more accurate flood simulation

Researchers at ORNL used TRITON to simulate the inundation of floods in Houston, Texas, and surrounding areas that resulted from Hurricane Harvey in 2017. Light purple indicates shallow water, while dark purple indicates deep water. Credit: Sudershan Gangrade / ORNL

A team of researchers from the Department of Energy’s Oak Ridge National Laboratory and Tennessee Tech University has created an open-source 2D flood model designed for a multi-architecture computing system. The Operational Needs 2D Runoff Immersion Toolkit, or TRITON, can use multiple GPUs, or GPUs, to model floods faster and more accurately than existing tools.

Flood modeling is an essential part of emergency preparedness and response. However, models must be fast and accurate – simulation results appear within minutes – to be useful tools for decision-making and planning. The higher the accuracy of the model, the greater the computational power needed to run, so organizations may resort to simpler models that sacrifice accuracy for the sake of speed. The computational power of GPUs allows high-resolution models to run calculations faster than simpler models that use only CPUs.

As high-performance computing has grown into an indispensable tool for science, it has also become a requirement for modern flood models to harness the power of hybrid CPU + GPU architectures. TRITON, whose development was funded by the Air Force Digital Weather Modeling Program, is specifically optimized for the multi-engineered design of supercomputers such as the IBM AC922 Summit at the Command Computing Facility in Oak Ridge.

Shih-Chieh said, “The unique thing about TRITON is not only that it uses GPUs — it’s not the only GPU-accessible Flood model. But it’s meant to use multiple GPUs simultaneously, Which makes it suitable for solving flood problems at the top.” Kao, ORNL Group Leader who led the project.

Water test. Credit: Oak Ridge National Laboratory

The team put the model through its paces on top to demonstrate its consistency, stability and some unique capabilities, such as runoff hydrograph. This optional data allows TRITON to simulate heavy floods – that is, local flash floods – as well as river floods. During a river flood, a stream or river swells and floods the floodplain. Using a 100-year FEMA data set as a benchmark, simulations using a runoff chart were more accurate than the basic hydraulic model alone.

“In order to truly understand the impact of floods, we need to understand flooding, which involves how deep a river is and the interpretation of different flood events: river flooding and flash flooding. Conventional flood models typically only address river flooding. TRITON can address both and provide more information on flood impact” Kao said. “If you have this dumping information, you can overlay it on the assets and assess which are at risk and which are not.”

In another test case, the team simulated a 2017 flood in the Houston metropolitan area caused by Hurricane Harvey. The simulation spanned 10 days and was built on two different hardware configurations: one that uses multiple CPUs and one that uses multiple GPUs. The results properly demonstrated the advantage of the flood model designed to run on a multi-GPU configuration. Even the smallest hardware configuration — one compute node with six GPUs — completed the simulation faster than the most powerful 64-node multi-CPU configuration.

As an open source toolkit, TRITON is freely available and can be used on a range of computing platforms – from laptops and desktops to supercomputers. Members of the research team are constantly developing new features and working on algorithms to extend existing capabilities to the operational level.

“TRITON will be a foundation for us to keep building on it, and we call it a toolkit for a reason. We keep building to make it more useful – that’s our vision. With more computing power, and lower prices, eventually everyone should have greater potential to use these capabilities to better simulate floods. “.

The study explores uncertainties in flood risk estimates

more information:
M. Morales-Hernández et al, TRITON: an open-source, multi-graphic, open-source 2D hydrodynamic flood model, Environmental modeling and software (2021). DOI: 10.1016 / j.envsoft.2021.105034

Submitted by Oak Ridge National Laboratory

the quote: New model harnesses supercomputing power for more accurate flood simulation (2022, July 28) Retrieved on August 6, 2022 from

This document is subject to copyright. Notwithstanding any fair dealing for the purpose of private study or research, no part may be reproduced without written permission. The content is provided for informational purposes only.

#model #harnesses #supercomputing #power #accurate #flood #simulation

Leave a Comment

Your email address will not be published.