In space, it’s almost always raining dust. Most of that dust is so small a microscope would have a hard time seeing it. Created by asteroid impacts, millions of these fine dust particles collide with Earth’s upper atmosphere every second. When they hit that atmosphere, they start a complex dance of plasmas and energy that can be difficult to see and understand.
Simulating that complex dance would allow scientists to understand what exactly is going on in the upper atmosphere. Still, so far, the complexities of the dance have confounded attempts to model them. Until now – a team consisting of members from John Hopkins University and Boston University used a supercomputer known as Stampede2 at the University of Texas to model what exactly happens to meteors when they hit the night sky.
Image of the Stampede2 supercomputer used in the study.
Credit – TACC
What happens to them might be hard to see with the naked eye. Dust particles light up the sky when looked at in the radar spectrum. When they hit the atmosphere, the particles go through a process called “ablation,” where they turn into glowing plasma, freeing electronics from their atomic bonding and creating a streak of light in the sky visible to radar telescopes.
Those telescopes can then track what direction the particle came from and how big it was, depending on the speed, trajectory, and length of time it was lit up. In addition, the actual spectra of the plasma itself could hold clues to the makeup of the meteor itself.
Larger meteors causing the visible phenomena of “shooting stars”.
Credit – Jacek Halicki
The data points hold clues to the meteors themselves and the composition and dynamics of the upper atmosphere. Scientists can bounce LIDAR signals off meteors to determine the upper atmosphere’s temperature, density, and wind speed. In addition, they can track wind direction by watching the plasma blow away, even if they only last for a fraction of a second.
But all this is extremely difficult computationally, and trying to understand what scientists are seeing would require a model to compare against. That’s where the new research comes in. Published in the Journal of Geophysical Research, the study utilized the Extreme Science and Engineering Discovery Environment (XSEDE)’s Stampede2 supercomputer to model three different types of simulations that feed into the model of meteors.
Meteor showers are more impressive versions of the type of plasma events modeled in the paper.
Those models essentially center around explosions, which are notoriously difficult to model, especially for heterogeneous objects like meteors. Like all engineering problems, Dr. Meers Oppenheim, co-author of the paper, tried to break it down into more manageable steps. The first of these was to model the molecule dynamics of the meteor’s breakup. In other words, how to model what happened to the individual meteor atoms when they are confronted with air molecules while traveling over 50 kilometers a second.
After that initial contact, the following simulation focuses on what happens next to the molecules. In particular, it tries to simulate where they fly to, at what speed when / if they become plasmatized. The third simulation uses a virtual form of radar to study the plasma to emulate what real-world radar systems would see.