Imagine the Universe!
rosat header image (

SNR and Thermal Conduction

Interesting Observations of a Class of SNR

An interesting case has come up. It turns out that many SNRs actually give off more thermal X-ray light from their interiors than from their edges (where we would expect shock-heated gas to emit lots of x-rays). This is just the opposite of what was predicted from theory. When observations run counter to predictions, it is natural for scientists to ask what could be wrong with or different from our model to account for the observations. In order to answer this question, a couple of years ago, Jeonghee Rho and Rob Petre from the Lab for High Energy Astrophysics used X-ray detectors to look at many of these SNRs. They created a catalog of their properties. Jeonghee Rho called them M-type SNRs for mixed morphology SNRs. At a recent conference on SNRs, they were called thermal composites. In this writing, they will be "thermal composites".


SNR IC443 - a thermal composite

"Why are these SNRs so different from the rest?"

Many ideas have been suggested. One idea is that the SNR expands past gas clouds in the interstellar medium (the gas that fills the galaxy). Once inside the hot SNR, the gas clouds are heated up and emit X-rays. Another possible cause of the extra X-ray emission from the center might be that while expanding, the SNR bumped into a big molecular cloud, causing a shock wave to bounce back into the SNR (like when you make a ripple of water in your bathtub and the ripple bounces off of the side of the tub). The edge of the SNR wave is where material piles up and emits X-rays, so if the wave has bounced back, it will cause the SNR to give off more X-rays from the interior than otherwise.

A third possibility, involving thermal conduction, is being explored by Robin Shelton. This possibility involves applying a concept familiar here on Earth, the conduction of heat, to a new situation. When heat moves from one place to another, like from near a home radiator to the other side of the room, we call that thermal conduction. Because the gas atoms near the home radiator are hot, they move quickly. When they bounce into other atoms, they make the other atoms move quickly. These second atoms hit third atoms, and so on, and so on. Eventually, many of the atoms in the room have been hit and sped up, so now all of the atoms are warm. That is how thermal conduction works here on earth.

Thermal conduction should be working in space, too, but there is a catch. Some of the atoms in space have been ionized, meaning that they have lost an electron or even several electrons. Both the ions (ionized atoms) and the electrons can cause thermal conduction. The catch is that if there is a magnetic field, then the ions and the electrons won't travel in straight lines, rather their paths bend because of the magnetic field. If the magnetic field is very strong, the ions and electrons bend so sharply as to almost make a circle.

Like a dog tied to a rope looped around a stake stuck in the ground, the ions and electrons just run in circles. They don't get to explore other parts of the neighborhood. They don't get to meet (or bounce into) the other dogs (ions and electrons) in the neighborhood. So, they don't cause the other ions and electrons to heat up. On the other hand, magnetic field have a direction and if that direction is the same as the direction in which the ions and electrons are moving, then the ions and electrons aren't slowed down by the magnetic field at all.

Although the situation is very complex, it is possible that thermal conduction has moved heat from the very center of the SNR to areas closer to the edge. This would make the center cooler than if there were no thermal conduction. Cool is a relative word -- the gas is still over a million degrees Fahrenheit! Because of a physical relation called the ideal gas law (pressure = density × temperature × a constant), this results in a larger amount of mass being in the center of the SNR (where the gas is cooler) than there would be if there weren't any thermal conduction. Because there is more mass, the center of the SNR will emit more X-rays. That is the idea that Dr. Shelton and her collaborators have been working on.

Because each of the thermal composite SNRs is different from the other thermal composites, it is possible that all of these physical processes (cloud evaporation, reflected shocks, thermal conduction) influence each thermal composite SNR, though which dominates can vary. Nonetheless, there is interest in finding out if the mechanism of thermal conduction could be used to explain some of the simpler thermal composite SNRs (the spherical cows of astrophysics. The idea is not to worry about some of the funny features of individual SNRs, but instead to try and figure out if this physical mechanism might be responsible for a general class of SNRs.

Why Do We Want to Know?

Why is it important to understand this particular class of remnants? Well, this is important for understanding SNRs and how a class of them evolves. But, it is also important for understanding the galaxy. The effects of thermal conduction on the interstellar medium haven't been studied much and could greatly modify some regions of our Galaxy, changing the densities or temperatures of the gas that makes up these regions and affecting such processes as star formation occurring within. SNRs can be used as examples or test laboratories, to study how thermal conduction works there, and then the knowledge can be extrapolated to the interstellar medium in general.

Observational Clues: Current and Future Work

How do we figure out if this model of thermal conduction in SNRs is correct? We use observational data from the ROSAT detector (, the Einstein detector, from radio continuum frequency detectors, 21 cm radio detectors (to measure the emission from hydrogen atoms), and even optical telescopes. We use very detailed computer programs to simulate the growth of a SNR for various conditions. We compare the X-ray and radio emission expected from the simulated SNR with that observed. If they don't match, then we say that something is wrong, either with our hypothesis or with the particular choice of parameters such as the supernova explosion energy, interstellar medium density, SNR age. If we try all reasonable combinations of the supernova explosion energy, interstellar medium density and SNR age and still don't get a good match, then we can say that we have tried the idea and it failed (or that the computer simulation wasn't complicated enough to simulate every important aspect of the SNR). In our case, we have found that the predictions agree reasonable well with the observations of our sample SNR, so we think that thermal conduction may be occurring and may be very important in SNRs.


Thank you to Robin Shelton for contributing to this article.

Imagine the Universe is a service of the High Energy Astrophysics Science Archive Research Center (HEASARC), Dr. Alan Smale (Director), within the Astrophysics Science Division (ASD) at NASA's Goddard Space Flight Center.

The Imagine Team
Acting Project Leader: Dr. Barbara Mattson
All material on this site has been created and updated between 1997-2012.

DVD Table of Contents
Educator's Index