Turbulence in the Interstellar Medium

[Note: I've attempted to write this document at an intermediate level. It is not, I hope, at the level of a technical journal article. However, one may need to know a few basic astronomy facts to understand it completely.]

See also my discussion of searches for extraterrestrial intelligence.

Turbulence in the Interstellar Medium

The space between the stars is not empty. Rather it is filled with a gas of varying density and temperature. By terrestrial standards this interstellar medium (ISM) is a vacuum. By Galactic standards, the ISM is vitally important as it is the raw material from which stars are formed.

Some of the ISM is ionized or in plasma form (i.e., the gas is hot enough that the electrons are stripped from the atoms, like the gas in the tubes of a neon sign). Presumably this plasma is formed by massive amounts of energy being injected into the ISM, for instance by the winds produced by the hottest stars, O and B stars, and near supernovae (the explosions which occur when O and B stars run out of fuel and collapse). By studying how stars interact with the ISM, we may gain a better understanding of how they are formed and how they die. Furthermore, radio astronomical observations must be conducted while looking through this plasma, so we must understand how it can affect our observations.

Scintillation

Scintillation (intensity fluctuations) results from the passage of a light wave through a random medium (i.e., one in which the refractive index varies randomly). A common example of scintillation is the twinkling of starlight, resulting from visible light passing through the Earth's atmosphere. The Earth's atmosphere is a random medium because weather systems keep it "stirred up"; a typical scale on which the atmosphere varies is a few to 10 cm. Other common random media include the air over hot pavement and the exhaust of a jet airliner.

Scintillation is also observed at radio wavelengths during observations of pulsars and extragalactic radio sources. In this case the random medium is the plasma component of the ISM and a typical scale is 1 AU (i.e., the Earth-Sun distance). The energy deposited by stellar winds from O and B stars and from supernovae are probably what keep the ISM plasma "stirred up." Like the case of a twinkling star, radio scintillation has the possibility of corrupting radio astronomical observations (after all, we don't see twinkling stars as they really are, but as they appear after their light has passed through the Earth's atmosphere) or possibly even hiding sources.

I've written a program which simulates this process.

[The word scintillate is derived from the Latin, scintillo -are, which means to sparkle, glitter. Cassell's Latin Dictionary]

Galactic Center

At or near the Galactic center is a very interesting source known as Sgr A* (the Galactic center is seen in the constellation Sagittarius, abbreviated Sgr). This source is very compact, with a size not too much bigger than the Earth-Sun distance. Yet, at radio wavelengths, it is one of the brightest, compact sources in the Galaxy.

Sgr A* (and some other compact radio sources in the vicinity) is angularly broadened, quite heavily. Angular broadening, like scintillation, occurs when light passes through a random medium. As the term suggests, angular broadening means that the observed size of Sgr A* is larger than its actual size. In one sense, then, our view of Sgr A* is corrupted by having to view it through the material responsible for the angular broadening. In another sense, though, this angular broadening allows us to study the ISM.

How Close to the Galactic Center is the Scattering Region?

Schematic of Angular Broadening Toward the Galactic Center
One key question about the angular broadening toward Sgr A* is, Where does it occur? All previous observations have not been able to constrain where the broadening material is along the line of sight. The material could be very close to Sgr A* (which would be quite interesting, because then it would be telling us about the Galactic center) or it could be some distance away from Sgr A* (which, while still interesting, wouldn't tell us anything about the Galactic center). The figure attempts to illustrate the differences. [It is worth noting that, although the scattering screen is shown as dark in the figure above, in real life it is not opaque. We can see right through it, it merely corrupts our view of what's on the other side. I've just made it dark in the image above to differentiate the scattering region from its surroundings.]

As part of my thesis, I made a set of observations looking toward the Galactic center attempting to detect other galaxies shining through our own Galactic center. Close to Sgr A* I don't find as many as one might expect. This lack of sources is consistent with the idea that the scattering region is close to the Galactic center, within about 150 pc (500 light-years). With some later observations, I found that one of the sources I studied was, in fact, an extragalactic source. I have argued that this source indicates the Galactic center scattering region is patchy, that is, it has holes in it. If an extragalactic source shines through one of these holes, we can see it. If it doesn't, it is so broadened that it effectively disppears into the background. A patchy scattering region is also consistent with some of the other Galactic center sources. The magnitude of broadening toward other Galactic center sources changes dramatically over small (angular) distances.

The observations and analysis formed Chapter 2 in my thesis. They have since been summarized in three papers:

GC Pulsars

A recent project involves looking for GC pulsars. The idea is that there should be beaucoup pulsars in the GC, as a result of both normal star formation and any star bursts. However, the same scattering that causes Sgr A* to be broadened will cause pulses from pulsars to be heavily smeared out. An easy way to understand this pulse broadening is to consider time delays. Consider a short pulse (short in time). The radiation aimed at us will travel directly to us; some of the radiation not aimed at us initially will be deflected toward us by scattering. This deflected radiation will take a longer time to get to us. The result will be that the initially short pulse will be smeared out.

Toward the Galactic center, pulse broadening may be as large as 500 seconds. Since typical pulse periods are 1 second, pulse broadening in the Galactic center may cause pulsars' pulses to become so smeared that pulsars will no longer appear to pulse. However, the fact that we can see Sgr A* (and other GC sources) suggests that if GC pulsars are bright enough, we should be able to detect them.

I'm currently in the middle of a survey for Galactic center pulsars. I'm nearly done and have found over 150 sources. Not all of these are good pulsar candidates, of course, but I hope to have a handful of good candidates when I'm done.

Anticenter

Toward the Galactic anticenter (i.e., 180 degrees away from the Galactic center), I have also observed extragalactic radio sources and attempted to measure their angular broadening. Here the objective is to determine how far out the plasma component of our Galaxy extends.

0526+245 [7 kB]  0558+232 [6 kB]
These images show two of the sources I've observed. The source 0526+245 appears to be a double, while 0558+232 looks like a point source. Note the scale on these images. One milliarcsecond (mas) is about the size of a dime in New York City as seen from Los Angeles. Impressive, no? The images were obtained from the VLBA.

I observed nearly a dozen sources. I found was that the no source showed huge amounts of angular broadening. My results were consistent with the turbulent ionized medium extending out only to 20 kpc (60,000 light years). I emphasize turbulent, because these measurements are only sensitive to a turbulent ionized medium. A quiescent ionized medium could extend to much larger distances.

These observations and analysis formed Chapter 3 in my thesis. They have since been summarized in two papers:

T. Joseph W. Lazio / rsd.nrl.navy.mil
Last modified: Sat Dec 1 13:33:52 2001
Best Viewed With ANY Browser