Stardust and What To Do About It

September 28, 2014

Astronomy, Science

DustBuster2aWhat is the nature of interstellar dust? How does it affect the starlight that passes through it? What can be done to locate and catalog the numerous, small, tenuous dust clouds that populate our galaxy at the higher latitudes? Is it possible to somehow vacuum under my refrigerator without moving the damn thing?

The nature and distribution of dust are issues we seek to understand. To that end we will need tools to help locate the clouds, and measure the effect of their presence. In this discussion I examine the nature of stardust and present some of the research being conducted on the topic of high-latitude galactic dust clouds.

When you go to an astronomy conference, the folks can be roughly categorized into two groups. There are those who are there to present hot sexy topics like black holes, dark matter, gamma bursts and so on. Then there are the lowly dust busters. Astronomers who study dust tend to sit together apart from those cool kids and talk among themselves about their calculators. But the fancy folks with the fancy headline grabbers depend on the dust busters big time, as the BICEP2 team recently confessed.

All the data collected by astronomers who study things far away have to be adjusted for whatever dirty little filter the galaxy has placed between here and there. Exactly what the dust is doing to the data depends on which way you look up. Just how much dust is that way? How dense? What are the shape of the grains? What are they made of? Are they all lined up magnetic-like over that way? The BICEP2 team initially announced they had found swirly goo in the cosmic poo but hey, dust does exactly that sometimes.


Dust is a relative newcomer to the Universe. Mostly it joined the gases only after the first stars of the cosmos had matured and their more over-bloated members exploded. It was these “supernovae” that seeded the interstellar medium with heavier elements, of which dust, by definition, is composed. Along with the introduction of dust came large aggregates of dust, which are the asteroids, moons, and rocky earth-like planets. But a lot of the dust still swirls and eddies along with gases in great big clouds called nebulae. Large collections of these nebulae orbit along with associated stars around common central points. These collectives are of course, the galaxies. To be fair, there was likely also some dust created early on by quasars spewing forth atoms at high speed from young galactic cores. Those high speed atoms crashed-smashed into other atoms which sounds like a good way to wind up with a bunch of hard little nuggets so we can’t blame ALL of it on supernovae. Just most of it.

The dustiest unkempt area of the Milky Way Galaxy is along the galactic plane, in which our solar system is firmly embedded. It’s a dirty gritty urban construction project area. Meanwhile, out away from the plane of the disk, both above and below, there appear to be isolated clouds of gas and dust scattered about, fluttering around quietly in the void. These high latitude dust bunnies are difficult to detect. They tend to be too cold to glow in the dark much. They are typically only 15 to 17 degrees above absolute zero which is freakin’ cold, but at least it is warmer than the background radiation of space, which is about 3 degrees Kelvin.

horsehead nebula

Galactic clouds of dust and gas can be damn fun to look at through telescopes, if you have nothing else lined up to do and the sky is clear and the moon is new and you have access to a big ass scope. The main targets are mostly the warmer clouds that do have some glow. Clouds that glow with a soft red light of their own are called emission nebulae. They are mostly hydrogen gas, warmed by young stars born within them. Another type of clouds, reflection nebulae, act like mirrors. These are mostly made of dust, scattering the bluish light from nearby stars back to us. Then there are the dark nebulae. The dark cold clouds don’t glow for shit, and aren’t even scattering a detectable amount of light from nearby stars. Thus they are the most elusive clouds to find and observe. Some of them can be found silhouetted, outlined by some brighter clouds that lay in the distance behind them. The most famous back-lit dark cloud is the Horse Head Nebula.

Amateur astronomers who swear to having seen the Horse Head Nebula are members of the liars club. I’ve seen it a couple of times. Seeing it is so difficult that, except under the most perfect conditions, it is impossible. So much is required to see the faint ghostly image; large optics (18 or more inches helps), transparent skies, still air, a high quality eyepiece, a special filter (O-III), good dark adapted eyes, and not least – an experienced observer. If all that fails, just say “oh wow” and let the next person in line have a look. Chances are they will do the same.

SONY DSCDetecting dust clouds doesn’t have to be left to the liar’s club. The discrepancy between a star’s intrinsic (real) color and its apparent (from Earth) color can be calculated, thus betraying the presence and density of an intervening dust cloud. That’s because dust will “redden” the light, that’s right, make it more red. Dust grains are small like the wavelength of blue light so dust scatters the blue away, much like the air in our atmosphere scatters the blue part of sunlight, and makes our atmosphere appear blue.

At high galactic latitudes (above and below the disk) the dust clouds are small and widely scattered and hard to map. Making calculations for hundreds of thousands of stars across a vast expanse of sky has always been the standard method for finding cold dust and that’s a rather daunting task. One solution is to use general models that estimate the extinction of starlight for a given set of galactic coordinates. This raises the issue, what degree of confidence can one have in a particular estimating model? Also, perhaps a more modern, non-labor-intensive method can be developed to find and catalog the dust clouds.

kansasDust clouds cause a certain amount of extinction of the starlight that passes through them, especially at the higher wavelengths. Put another way, interstellar dust particles extinguish light more efficiently at short wavelengths (blue) than at long wavelengths (red). Having attenuated the light more in the blue region than the red, we say that the cloud has “reddened” the starlight. Stated succinctly, the extinction coefficient (Ql) goes as the radius of the dust grains and inversely as the wavelength of the incident light, Ql ~ r/l. There are exceptions at certain wavelengths due to the atomic composition of the dust and resonance, but the mean result is that the color (surface temperature) of stars can appear redder (cooler) than their actual value. The discrepancy can be detected by considering the absorption/emission lines in the star’s spectrum (Carroll & Ostlie p. 440).

At one point in my twisty turny life-path, I undertook an effort to better account for the location and affects of galactic dust. The goal of my project was a comparison between photo-metric extinction calculated for lots of individual stars, and the corresponding estimated extinction maps. Data from the Bright Star Catalog was used to calculate the extinctions. Estimating was done using a widely accepted model (Burstein & Heiles 1992). My claim is that the comparison allows for an evaluation of the level of confidence one can have in the estimated extinction model.

Okay here’s how it works. Certain stellar values from the Bright Star Catalog (Hoffleit & Warren 1991) were downloaded from the Internet using the Astronomical Data Catalog Viewer from NASA. The total sample included 3,802 stars.

Here are the types of data for each star that was downloaded from NASA:

• galactic longitude (l)
• galactic latitude (b)
• visual magnitude (V)
• apparent color (B-V)
• spectral type and class (SpType)

dusty-caseHaving acquired the data, I developed a software application called “E-calc” thanks to MASSIVE help from my buddy Huw Upshall of Microsoft, whom I plied with promises of beautiful ale and fresh women etc. The purpose of the app was to calculate the magnitude of extinction of all the target stars using the data download. The program calculates the extinction of starlight, i.e. the difference between observed magnitudes and intrinsic magnitudes, for luminosity class V stars at galactic latitudes greater than 20 degrees. The process uses a file that equates each spectral type with an intrinsic magnitude and color (B-V)o (Carroll & Ostlie append. E). The calibration file is included with the program, but is accessed separately so it can be modified by the user.

The input file includes galactic longitude and latitude (l,b), luminosity class,
observed color (B-V), and spectral type. At the end of the day the output file is designed to provide a calculated reddening for each star along with it’s location within the galaxy.

Output Format:
l, b, E(B-V)

The program filters the input such that certain stars are eliminated from the sample. The stars NOT eliminated meet the following criteria:

• Latitudes more than ± 20 degrees.
• Luminosity class V.
• Spectral types OBAFGKM.
• Only the stars with complete data records.

The essence of the calculation is simply to subtract the intrinsic color from the apparent color, producing the magnitude of extinction (reddening):

E(B-V) = B – V – (B-V)o

Another_one_bites_the_dustHigh-latitude extinction estimate models, such as the Burstein & Heiles method, are often calculated using atomic hydrogen data in the 21-cm spectral line. The assumption here is that most dust is associated with the gas, and that most gas is atomic (H1) as opposed to molecular (H2). The data and code for the Burstein & Heiles model can be obtained from the NASA Astronomical Data Center. The problem is that the B&H method can miss significant amounts of dust associated with molecular hydrogen, which has an emission signature in the UV, inaccessible from Earth-based telescopes.

A total of 580 stars survived the E-Calc filter (high latitude class V stars). The calculated extinction values were then plotted against the estimated values. (Figure 1).

table_01The stars with the greatest calculated color excess, i.e. those with reddening greater than 0.2 magnitude, were singled out (Table 1).

The 19 high-extinction stars have been plotted against their estimated extinction. A solid line has also been plotted which represents unity (points where the calculation equals the estimate precisely) (Figure 2).

AHA! From the plot of the unity line, it is evident that all the calculated values for these stars are greater than the estimated values. From this sample, it appears that the estimated extinction model might represent only the lower limit to the actual extinction.


Next step: some possible explanations for anomalous reddening were investigated;

table_02• Distance from Earth.
• Galactic latitude.
• Proximity to known galactic clouds.

The distance in parsecs (dpc) for the 19 star sample was calculated using the stellar distance modulus:


Where V is the visual magnitude, MV is the intrinsic magnitude and
AV = 3EB-V.

Available data were used for MV values (Carroll & Ostlie A-13). The results are displayed in Table 2.

The calculated extinctions for these stars were plotted vs the calculated distances in parsecs (Figure 3).

With one possible exception there seems to be no correlation between extinction and distance. It would be unfair to attribute reddening to accumulative effect of rarefied interstellar dust over a great distance.

Okay so far so good. For the next investigation, we might expect less extinction at higher latitude since most of the dust in the galaxy is located close to the galactic plane. So the extinctions were plotted against latitude (Figure 4).
Again no correlation is observed. The absence of stars in the center is an artifact from the E-calc program filter that eliminates stars near the galactic plane, i.e. less than 20 degrees latitude. Other than this there is no clear pattern.

table_03Next, each of the sample stars was correlated with known high latitude galactic clouds in their proximity (APJ 447). The distance to the center of the nearest known cloud is given in degrees (Table 3).

Some stars, such as number 5, is a distant 6.7o from the associated cloud, but that cloud (MBM 18) is large, perhaps 10o on a side. Other examples, such as 15 and 16, are isolated from known clouds of dust and gas.

At last the fun part. A map of known high-latitude dust clouds, over-plotted with the reddened sample stars, was generated (figure 5).


With the sample stars plotted on a recent map of molecular gas clouds made from measuring the infrared excess over the interstellar medium (Reach), a rather interesting section of that plot is enlarged and displayed (Figure 6).


Stars 5, 6, 9, and 11 seem to have located the Taurus Star Cloud. Numbers 12 and 14 appear isolated but they are very near the plane of the Milky Way where dust is known to be plentiful.

The really interesting star is number 4. It seems to have located a lonely uncataloged high-latitude dust cloud. This is precisely the result that was hoped for. This is evidence indicating that the process employed may indeed find remote dust clouds at high galactic latitudes. That’s my story and I’m sticking to it.

To summarize the research, first the visual and intrinsic colors of 3,802 bright stars were subjected to processing in order to determine the degree of extinction suffered by the light on route to Earth. From the total sample 580 stars survived a filtering process and of these 19 stars showed reddening greater than 0.2 magnitude. This final sample was examined for correlation to the distance, latitude, and proximity to known galactic clouds.

A correlation with distance was considered because of the possible accumulative effect of diffuse dust throughout the interstellar medium. No such correlation was found in the sample. Reddening could not by attributed to distance.

A latitudinal correlation was considered to rule out a systematic and isotropic effect caused by our position embedded in the galactic plane. The results for the sample could not be explained by latitude, as no such correlation appeared.

Some correlation was found between a few of the sample stars and known high latitude galactic clouds such as Taurus, Orion, and Ophiuchus. In fact, the list of associated clouds reads like a who’s who of known high latitude clouds. This suggests that the procedure does indeed use extinction to find clouds. Six of the nineteen stars appear within three degrees of a known cloud. Thirteen are within eight degrees of a known cloud. Only two are greater than ten degrees.

The reddened stars of the sample not associated with known clouds can be regarded as evidence of unknown clouds of dust, high above the galactic plane. The effect of this dust on starlight is typically greater than estimates made by Burstein & Heiles. At best, we can consider the B&H estimate to be only a lower limit of the actual reddening of starlight by interstellar dust.

Understanding high-latitude extinction is important in extra-galactic astronomy. Toward that end we can use a process like the one outlined here. Thanks to personal computer power we can now quickly examine the extinction of starlight of a great number of stars, beyond those of the Bright Star Catalog, to better map the high latitude dust, and generally improve our extinction estimate models.

All of my involvement regarding this topic has been possible with undying thanks to my research adviser, dust buster extraordinaire, Dr. Kristen Larson of Western Washington University.

Carpe Noctem

Sources Cited & References:

Morrison, David, Sidney Wolff and Andrew Fraknoi.
Abell’s Exploration of the Universe 7th ed.
Philadelphia, PA. Saunders College Publishing. 1995.

Burstein, D. and C. Heiles. 1992.
Astronomical Journal, Vol. 87. p.1165.

Carroll, Bradley W., and Dale A. Ostlie.
Modern Astrophysics.
Reading, MA. Addison-Wesley Publishing Company Inc. 1996.

Magnan, L., D. Hartmann, and B.G. Speck.
Astrophysical Journal Supplement.
Vol. 106. p447. 1996

NASA Astronomical Data Center.
Hoffleit E.D., and W.H. Warren Jr.
Bright Star Catalogue 5th Revised. 1991.
The Astronomical Data Center Bright Star Catalogue. ID 5050.

Reach, William T, William F. Wall, and Nils Odegard
The Astrophysical Journal. 507:507-525. 1998 November 10.
The American Astronomical Society.

, , , , , , , ,


Subscribe to our RSS feed and social profiles to receive updates.

No comments yet.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: