Lecture 22: The Milky Way
Detailed Shape of the Milky Way
-- comparison of various views of the Milky Way with external galaxies suggests strongly that the Milky Way is a spiral galaxy
Milky Way as seen by the Cobe satellite in the near-infrared.
How the Milky Way might appear if viewed face-on:
-- mapping of some spiral arms is possible using OB, stars, HII regions, and HI 21-cm radio data, radio maps rely on our knowledge of the rotation of the Milky Way
Rotation of the Milky Way
-- plot of rotation speed versus distance from the center of the Milky Way reveals that the stars do not orbit the center of the galaxy like planets orbiting the Sun (Keplerian rotation) but rather follow a much flatter rotation curve.
Keplerian rotation curve:
Observed rotation curve:
This difference implies much more mass lying far from the Galactic Center than one would infer from stars and gas detectable at visible, IR, and radio wavelengths. This in turn implies that most of the mass (~90%) of the galaxy is invisible and in some as yet undiscovered form! This "missing" matter is dubbed dark matter. Other spiral galaxies have similar ratios of visible to invisible matter.
-- rotation of the Milky Way can be used to determine distances of HI clouds along the line-of-sight
l=Galactic longitude V=rotational velocity
R = distance from Galactic Center
VSun=Sun's velocity around the Galactic Center
RSun=Sun's distance from center
d=distance to HI cloud Vr=radial velocity of cloud
The observed radial velocity of the cloud will be the difference between the Sun's component of rotation and the cloud's along the line-of-sight:
The angle SPC =90°+ and from the law of sines,
The radial velocity is
From the law of cosines
where d= distance to the cloud.
For example, the distance to a cloud at l=135° with a radial velocity of -50 km/sec can be computed knowing the Vsun=220 km/sec and Rsun=8.5 kpc:
The flat rotation curve of the galaxy says that V~220 km/sec so
Using the law of cosines gives the distance:
By observing in many different directions, a map of cloud locations and hence spiral arms can be built up. However, clouds can have non-circular motions and so the map will not be perfect. Spiral arms determined from radio data and OB stars and HII regions agree fairly well nonetheless!
HR Diagrams for some Stellar Groups
The HR Diagrams for the nearest stars and the apparently brightest stars look quite different (nearby stars on the left and bright stars on the right).
Nearby stars ............................. Apparently brightest stars
If we plot histograms of the numbers of stars in each category, we discover that the commonest star is an M dwarf, less luminous than the Sun. The apparently brightest stars are actually mainly intrinsically very bright and rather rare in the galaxy.
(recall that the Sun's absolute magnitude is +4.7 to help in comparing these two plots). These differences result from 1) the initial mass function being weighted towards low mass stars and 2) high mass stars disappear realtively quickly while all the low mass stars are still on the main sequence.
Look at the trends of stellar properties with position in the Milky Way:
|Property||Population I||Intermediate||Population II|
|Metals (%)||3-4||0.4-2||0.4 or less|
|Total Mass (Msun)||2x109||5x1010||2x1010|
|Typical objects||Open clusters, HII regions, OB stars||Sun||Globular clusters, RR Lyrae stars|
Notice the systematic trends -- younger objects have thin, disk -like distributions and lots of metals while older objects have nearly spherical distributions and few metals.
The trend of metallicity versus age for the galaxy is quite strong and is a consequence of the build of metals via stellar evolution leading to supernovae and the incorporation of this material into the next generation of stars.
The trends of properties with spatial distribution suggests strongly a model for the formation of the galaxy which is similar to the model for the formation of the Solar System:
The Center of the Milky Way
Ever since black holes were suggested as the power sources for Active Galactic Nuclei (AGNs) such as Seyfert Galaxies and QSOs, we have speculated on whether the center of our galaxy might contain a black hole.
Because the center of the Milky Way is by far the closest galaxy nucleus, we can study details that will remain indistinguishable in other galaxies for a long time.
"Galactic Center" here will mean the central ~10 parsecs of the Galaxy.
Sampling of phenomena observed at the Galactic Center:
1) the stellar population including evidence for star formation there in the last 50 million years or even less.
2) interstellar material including both ionized gas (HII regions) and molecular clouds which orbit the Center in a ring with an inner radius of about 2 pc. Hot dust is also observed.
Left is hot dust, right is radio map of ionized gas:
3) strong magnetic fields (milliGauss) as compared to elsewhere in the Galaxy
4) a compact radio source called SgrA* which is quite unlike any another radio source in the Galaxy.
5) radial velocities and proper motions of both stars and gas which imply the existence of a large, unseen, compact object. Large means a mass=~2.5x106MSun.
The discovery that the radio source SgrA* corresponds to the dynamical center of the Milky Way and coincides with the large, dark mass has lead to the realization that SgrA* is a black hole, albeit a puzzling one:
-- why no accretion disk in spite of an abundance of nearby interstellar material with angular momentum?
Stars at the Center of the Milky Way
--- 30 years ago the center of a spiral galaxy was presumed to consist of an old stellar population with no on-going star formation. This was based on the properties of the bulges of spiral galaxies which are predominately Pop. II.
--- Infrared studies revealed the presence of young stars in our own Galactic Center as well as "bursts" of star formation in the nuclei of other spirals. Radio data showed the existence of abundant molecular gas.
--- Detailed study of our Galactic Center has been hampered by the large amount of interstellar dust along the line-of-sight. The extinction at visible wavelengths (at V) amounts to 30 magnitudes! Only by observing at longer wavelengths such as 2200 nm = 2.2µm (K filter) is the extinction low enough to observe anything.
--- Infrared spectroscopy has shown that central parsec has a large number of blue supergiants and some red supergiants. Because of uncertainties in the models for these stars, the exact age of these stars is rather uncertain but cannot be more than 50 million years and could be a lot less.
To pin down this age more accurately, an observing program using the NICMOS camera on HST was carried out. By searching for the tip of the main sequence from the last episode of star formation, the age of the young population could be determined.
--- need to observe in the infrared because of the heavy extinction (but infrared not the best wavelengths to distinguish the properties of relatively hot stars)
--- need to observe using the Hubble Space Telescope because of extreme crowding of stars
Can we estimate what brightness level and accuracy we must achieve with HST to detect the tip of the main sequence? (note that J is the letter assigned to 1.25µm, H is for 1.6µm, and K is for 2.2µm).
Relevant facts: d=8.5 kpc for the Galactic Center
Age ~ 10-50x106 yrs
AK = 3.3 magnitudes
Intrinsic colors of relevant stars:
|H-K||Abs. mag at K, MK||Tº K||Main Seq. Lifetime|
Clearly to distinguish O6V stars from O9.5V stars requires very accurate J-K colors and can't be separated on the basis of H-K colors. The brightness will help distinguish these stars.
Why aren't the infrared colors better at telling such hot stars apart from one another?
-- because the peak wavelength for these stars is at 5.1x106/T=5.1x106 nmºK/30000ºK= 170nm, a much shorter wavelength than even J at 1.25µm=1250nm. Blackbodies at wavelengths much longer than their peaks all asymptote to the same form so the infrared won't identify these stars easily.
An additional complication comes from the heavy extinction which is not identically the same for each star.
Measuring two colors would be enough to produce a color-magnitude diagram analogous to V vs B-V (say K versus J-K), but to separate the intrinsic colors of the stars from the reddening by interstellar dust will require additional colors.
What I actually did:
1) Observed the Galactic Center using 4 filters.
2) Observed some nearby regions for comparison to see if the Center has any star indicative of the tip of the main sequence that are not see elsewhere near the Center.
Step 2 is required because our line-of-sight to the Galactic Center passes through so much of the galaxy that there is the possibility of seeing dimmer foreground stars that could mimic the properties of more distant stars but intrinsically more luminous stars. The distribution of previously known blue and red supergiants indicates that the star formation at the Galactic Center is confined to a small region .
What I've Found So Far:
Can we be certain that there is a black hole at the Galactic Center?
Recall the standard tests for the existence of a black hole:
Test 1 has so far yielded only evidence for a very small accretion disk at the Galactic Center. Test 2 has provided some very compelling evidence:
Consider a star whose proper motion has been measured to be equivalent to 1000 km/sec and which lies only .01 pc from SgrA*:
Recalling that MSun=2x1030kg yields M=2.3x106Msun. This mass is located at Sgr A*. How compact is SgrA*?
VLBA measurements of SgrA* set limits of ~3AU for the size of SgrA*:
So the radio data are probing done to a distance of less than 70 RSchwarzschild so it is very difficult to postulate any form of matter other than a black hole to explain SgrA*.
Open Issues in Galactic Center Research
No longer any doubt that stars have formed recently but how? The inner edge of the molecular ring is defined by the Roche limit for a reasonably dense gas cloud near a mass concentration like the black hole (or whatever the dark object is). This constraint plus the strong magnetic field at the Center would seem to make cloud collapse very difficult.
At what point will any model of a black hole become untenable? Will NICMOS push the detection limit for any flux so low that no model will work?