Significant work in the area of tornado simulation has been conducted by a team of University of Oklahoma researchers with the aid of the Texas Advanced Computing Center's Ranger supercomputer. The team, led by professor Ming Xue, simulated several real-world tornadoes with remarkable accuracy, and the analysis of the model output is providing clues about the formation of tornadoes and the role of microphysical processes within the cloud. The difference in scale between tornadoes and the thunderstorms that generate them, combined with microphysics inside the cloud and the Doppler radar observations that must be assimilated to permit forward prediction, make powerful supercomputers a necessity for modelers. The simulations that Xue's group achieved used new techniques for assimilating high-resolution Doppler radar data and for blending this information into three-dimensional visualizations of evolving storms.
"Our theory says that the cloud microphysics affect a thunderstorm's cold pool, and the cold pool affects how the mid-level updraft and rotation are positioned relative to the low-level rotation," Xue says. "We believe this relative position is a key factor affecting whether a thunderstorm can produce a tornado or not." However, Xue says it will be some time before a tool for forecasting tornadoes accurately and in real time can be developed from these methods. This advance will require next-generation high performance computing systems capable of computing at the multi-petascale level.
From The University of Texas at AustinView Full Article