NCAR animation shows multiple views of 2014 storm’s behavior
Oct 2, 2015 - by Staff
Oct 2, 2015 - by Staff
October 2, 2015 | A remarkably detailed visualization of last year’s Hurricane Odile is helping scientists better understand major storms. NCAR employed the kind of 3D software used in Hollywood movies to provide multiple perspectives within the storm, including the complex interaction between the hurricane and steep terrain.
"This new type of visualization shows a storm’s structure in ways that are more natural for the human mind to visualize," said NCAR scientist Jonathan Vigh, who helped coordinate the project. "The result is a stunning new look into the workings of a hurricane."
Hurricane Odile reached Category 4 intensity with 140-mph winds, weakening slightly before it swept across the Baja California Peninsula in northwestern Mexico a year ago. The September 2014 storm caused widespread damage, severe flooding, and power outages. Mainland Mexico and the U.S. Southwest also were affected, with 15 deaths blamed on the storm.
A scientific visualization based on storm observations, the animation was created by NCAR software engineers Tim Scheitlin and Matt Rehme. Scheitlin said the project provided an opportunity to integrate Autodesk Maya, a 3D software that has dominated the Oscars (from Finding Nemo to Harry Potter to Frozen), with other visualization tools. The result is movie-quality renderings of the storm’s encounter with the peninsula.
Multiple variables are depicted, including precipitation, wind speed, relative humidity, and vorticity – a measure of the spinning motion of the storm.
Odile spun up quickly at first. Then the peninsula, with mountains rising as high as 10,000 feet, provided friction to slow it. Odile’s rotational (angular) momentum was so depleted that a separate, opposite spin occurred in the upper regions of the storm.
Toward the end of the animation, the influence of the mountains becomes apparent, with the air dramatically drying as the storm flows downstream of the mountain ridges.
The animation is based on a Hurricane Weather Research Forecast simulation of Odile. The HWRF system is a high-resolution atmosphere-ocean model that is customized to simulate hurricanes for weather forecasting.
Vigh said it would be useful to scientists to see 3D visualizations of other hurricanes. "This really brings the structures of a storm forefront – features that can be conceptualized but difficult to see."
A lot of complex interactions occur inside a hurricane. These storms also interact with seemingly remote weather systems. A system that eventually influences the hurricane’s path may be thousands of miles away from the storm at the beginning of the forecast. Therefore, hurricane models must represent a large region while also depicting the small details within individual thunderstorms.
But even today’s massive supercomputers aren’t fast enough to rapidly compute an ultra-high resolution forecast over such a vast region. To include enough data, the Hurricane Weather Research and Forecasting model (HWRF) divides the region of study into three-dimensional grid boxes of varying resolutions, with key physical processes such as temperature, moisture and wind speed measured at each grid point.
The highest-resolution grid sits over the center of the hurricane, providing the detail needed to simulate individual thunderstorms. Grid points were less than 2 miles, or 3 km, apart to simulate storm activity inside Odile. To capture larger, more-distant weather systems, the points were spaced up to 16.8 miles, or 27 km, apart. The boxes extended 61 vertical levels, enabling scientists to see, for example, how the air dried as Odile came off the mountains of the Baja Peninsula.
Computational Modeling: Mrinal Biswas, NCAR Research Applications Laboratory/Developmental Testbed Center
Visualization and Postproduction: Tim Scheitlin and Matt Rehme, NCAR Computational and Information Systems Laboratory. Additional contributions by NCAR scientists David Gochis, Mary Haley, Dennis Shea, Richard Valent, and Jonathan Vigh.