Dynamics of Mesoscale Weather Systems

A particular emphasis in MMM has been high-impact convective weather systems, such as squall lines, hurricanes, and tornado-producing thunderstorms. In a recent series of studies, MMM scientists showed how numerical models could produce hurricanes with wind gusts greater than ever observed, despite very realistic initial conditions. They found that a particular combination of thermodynamics and flow structure in hurricane eyewalls acts as a form of atmospheric front in which mesoscale gradients are enhanced. Turbulent phenomena called “eyewall mesovortices” form in this region, and can limit the tendency to enhance gradients, but MMM scientists showed that extraordinarily small horizontal grid spacing (< 100 m) is needed to accurately simulate this process. Utilizing the Yellowstone supercomputing system, researchers have studied these eyewall mesovortices dynamics and their effects on the mesoscale structure of hurricanes. 

MMM scientists continue to explore and document the dynamics of numerically simulated mesovortices in idealized hurricanes, by evaluating model output against observational data collected and analyzed by colleagues at NOAA’s Hurricane Research Division. In one study scientists examined model output and found that the strongest wind gusts occur within shallow (< 1 km) mesovortices, which are coincident with strong updrafts along the eyewall of the simulated hurricanes (see figure below). A thorough analysis of the entire NOAA dropsonde database was conducted, and dropsondes with windspeed > 90 m/s were identified for future comparison with numerical model simulations.   

3D Visualization
Figure: Three-dimensional visualization of coherent structures associated with a strong (90 m/s) wind gust in a simulated hurricane. Purple shading illustrates strong updrafts (vertical velocity > +12 m/s), and green shading shows strong vertical vorticity (a region of rapidly rotating air; in this case, vertical vorticity > +0.15 s-1). Color shading at the bottom denotes wind-speed perturbation, showing +20 m/s perturbations from the mean flow.

In another study, advances in computing resources were utilized to simulate phenomena more accurately through higher resolution and more advanced physical parameterizations. As an example, several advanced subgrid turbulence parameterizations for large eddy simulation (LES) have been developed as part of this work. Two subgrid turbulence models have been added to the CM1 numerical model and evaluated as part of this project. The Nonlinear Backscatter and Anisotropy (NBA) model was found to have a minimal effect on results. However, the Two-Part subgrid model of Sullivan et al. (1994) improves near-surface profiles of wind speed; specifically, shear in the lowest 100 to 200 m above the ocean better matches observed profiles from dropsonde observations.   

MMM scientists continue to share the tools developed from studies of the dynamics of mesoscale weather systems. When possible and appropriate, lessons learned from simpler numerical models are transferred to WRF and/or MPAS.  Last year code was developed for this work and has been shared with a team of researchers at the University of Miami, who have an NSF grant to study the structure and dynamics of tornadoes with grid spacing < 10 m. 

MMM scientists, in collaboration with NOAA’s Hurricane Research Division (Dr. Sim Aberson), continued research on extreme winds in hurricanes, focused on the analysis of extreme wind gusts and updrafts in dropsonde data. NOAA has been particularly interested in this study because they conduct routine aircraft surveillance into hurricanes and are sometimes impacted by severe gusts and updrafts. Additionally, based on this work, several modifications were made to the CM1 numerical model to improve stability of the code at high wind speeds. CM1 is used by several university collaborators for a variety of applications, including research and education.