Slashdot Mirror
Milky Way Is Twice the Size We Thought
Peter writes to tell us about a research group at the University of Sydney in Australia, who in the middle of some calculation wanted to check the numbers everybody uses for the thickness of our galaxy at the core. Using data available freely on the Internet and analyzing it in a spreadsheet, they discovered in a matter of hours that the Milky Way is 12,000 light years thick, vs. the 6,000 that had been the consensus number for some time.
They're measuring the sea of electrons between the stars, which they assume stops at the 'edge' of the galaxy.
... If you know the distance to the pulsar accurately, then you can work out how dense the WIM is and where it stops - in other words where the Galaxy's edge is.
FTA:
"As light from these pulsars travels to us, it interacts with electrons scattered between the stars (the Warm Ionised Medium, or WIM), which slows the light down.
There are some who call me
Leave it to wikipedia to cite as a source a NASA edutainment page aimed at grade schoolers.
What the "source" doesn't mention (because it's not meant to give an in depth answer) is that the galaxy is ~1000ly thick on average. It is quite a bit thinner along its edge, and quite a bit thicker in the core.
It's pretty hard to measure the size and shape of the Milky Way simply because we are stuck in the middle of it. Measuring the size and shape of far away galaxies is a lot easier because we have a better view. Our galaxy is a flat disk with spiral arms where we are in one of those arms - the overall structure is very hard to measure from that perspective. To complicate things further there is quite a lot if interstellar dust that messes up our view in certain directions.
As an analogy - imagine being stuck in a traffic jam. Figuring out the extent of it is very hard from the view you get from your car. A helicopter in the sky has no problems though.
siener's youtube channel
Hyah you go.
By and large, language is a tool for concealing the truth. -- George Carlin
The pulsar data they downloaded from the internet largely did use big, expensive instruments—this work is a new, improved analysis of a large sample of already-published data from many sources.
They did not use the canonical space-averaged electron density for the WIM (0.03 cm-3); they used pulsars with independent distance measurements*. What's different about their work from previous estimates of the scale height of the WIM is that they did not use pulsars with any of several other distance measurement techniques that are less reliable. In particular, one of the commonly used distance measurement techniques uses absorption due to neutral hydrogen in the plane of the Milky Way. However, the neutral hydrogen (cold neutral medium, CNM) disk is considerably thinner than the WIM disk (scale height of 100–250 pc, depending upon whom you ask, versus 1000 (the old result)–1800 (their new result) pc for the WIM), so that technique only works at all well for pulsars in the plane (and is still model-dependent even then), which makes it a biased sample for measuring the height of the Milky Way's disk.
These authors also limits themselves to galactic latitudes |b| > 40 degrees, which means that they're sampling a relatively local cylinder about the Sun. Therefore, their sample isn't contaminated by spiral arms or many classical H II regions (gas ionized by hot, massive stars), which will change the result.
This result is a fairly dramatic revision of the scale height of this phase of the interstellar medium and, consequently, the weight of the medium. (In fact, it's the phase I make my living studying, so it's very important to me!) However, this does not have any bearing on the scale height of the stars (which contain 85% of the mass in the Galaxy) or the neutral hydrogen. It also doesn't change the total amount of ionized gas in the WIM. (That column density is measured very accurately by pulsar dispersion.)
The WIM is certainly not uniform throughout the Galaxy. It is a turbulent medium with varying densities, and it only fills ~20% (that number is highly uncertain, to a factor of two or more, I would say) of the volume within the 1000–1800 pc high disk. However, particularly over the path lengths the more distant pulsars sample, those local differences should be pretty well averaged out.
The discrepancy with previous work is largely due to a tremendous amount of progress in recent years measuring parallax distances to pulsars, largely using very long baseline interferometry. Distance measurements in astronomy are notoriously difficult, and improvements will continue for years to come.
* They relied only upon distance measurements determined in one of two ways: parallax (the only direct distance measurement method in astronomy, useful for relatively near pulsars—out to about 1000 pc=3000 ly, with decreasing accuracy further away) and association with globular clusters. Globular clusters contain thousands of stars that were formed at about the same time and have the same heavy element content, so their distance can be determined based on standard, well-known stellar evolution models and a color-magnitude diagram. These two distance measurements are about as accurate as a pulsar distance measurement will get in the foreseeable future, although particularly the parallax distances will continue to improve both in quantity and quality.