Planets and Disks Around Other Stars

If the formation of our own solar system from a disk of material is a natural consequence of star formation, then we ought to see many other stars with disks and planets. What do we actually see? As it happens, we can see stars and their “planetary” disks in various stages of formation. We can also detect the presence of planets orbiting other stars, although we have not yet been able to get a clear image of one.

Star and Disk Formation It had been suggested well over a century ago that stars were forming in “nebulae”, great clouds of glowing interstellar gas such as the Orion Nebula. This idea turned out to be correct, but it took a century for the idea to be truly confirmed.

		In the mid-to-late of the 20^th Century, a number of pieces of the puzzle began to come together. It was discovered that the youngest stars were in fact located in these nebular regions. These glowing clouds were also found to be adjacent to large dark clouds of molecular gas and dust. Streamers and blobs of glowing material were seen to emanate from these dark clouds, and radio spectra showed that these streamers were tracers of oppositely-directed molecular flows or jets.
	Theorists began to try to explain these flows as the result of the accretion and expulsion of material in a disk similar to what our Sun had when the planets were forming. Simply put, these were examples of what our own solar system was once like. Observations at infrared (IR) wavelengths are good for penetrating through the dust, and star-like objects were discovered at the heart of these flows.

Baby Stars A number of “baby” stars were also being found just outside the dark clouds. These stars are a little older than their embedded cousins, but are not full-fledged stars just yet. By looking at these in the IR , it was possible to detect excess IR emission from silicate materials located near these stars. Could these be the disks we were seeking?

Adult Stars

In 1983, the Infrared Astronomical Satellite (IRAS) made a survey of the sky at infrared wavelengths: 12, 25, 60, and 100 mm. If a star’s light were emitted entirely from is surface, the intensity (flux) should be described fairly accurately by a blackbody curve appropriate to its surface temperature. What IRAS found was that a rather significant (15% or more) of all “normal” stars also had IR excesses. One of these was the bright star Vega, and these stars have since been called “Vega-like” stars for want of a better term.

“Normal” stars	IR-excess stars

The IR excesses were indicative of lots of stuff with temperatures at a few hundred Kelvins located somewhere near these stars. It was known that many newly-forming stars possessed similar IR excesses, and these were due to dust. Was the emission from the older normal stars the remnants of their young disks, similar to the one the Sun once had?

Beta Pictoris

One of these “normal” stars had a tremendous IR excess. Using a coronagraph to block out the light of the star (and making any surrounding material visible), two astronomers (Brad Smith and Rich Terille) discovered that Beta Pic had a very substantial disk of dust. Spectra of the dust confirmed that it was composed at least in part by silicates.

Now the race was on to image more of these disks, and study them mineralogically.

Young Stars and Protostellar Disks

Using adaptive optics, coronagraphs, and the Hubble Space Telescope, we now have numerous examples of disk systems round young stars.

In one star, HD 163296, we see the weak remnants of a disk and the jets of material still being ejected from it.

Mineralogically, the dust in these systems is “familiar territory”. Observations of one of these, HD 100546, made by the Infrared Space Observatory (ISO) satellite a few years ago, showed spectral features identical to those seen in Comet Hale-Bopp using the same instrument! Most of the sharp spectral features are due to crystalline Mg-rich olivine (“Forsterite”). Some of the others are due to organic materials called PAHs (Polycyclic Aromatic Hydrocarbons). Interestingly, ultraviolet spectra of the stars indicates material falling toward the star which are most easily explained as star-grazing (or hitting) comets just like those we see still hitting the Sun today, but in much greater quantities. The easiest way to put the comets on star-grazing orbits is the same one that operates on our own solar system: gravitational tugs by planetary-mass objects.

All of this fits a consistent picture. Many stars form with a disk of gas and dust. As the disks gradually cool, the dust grains (silicates, ice, solid organics) settle into a plane and grow to great sizes. These coalesce to form planetesimals of rock and metal inside the snow line, and cometesimals of rock, metal, and ice outside the snow line. The planetesimals inside the snow line form the terrestrial planets and unaccumulated asteroids. The cometesimals outside the snow line form the cores of the jovian planets which, if they are massive enough an located in the densest part of the cold disk, also accumulate gaseous H₂ and He (i.e., like Jupiter and Saturn did in our own solar system).

If this is happening all over the place, we ought to see stars with fully-formed planets.

Planets Orbiting Other Stars

There are a number of possible ways to detect planets.

Imaging Planets shine primarily by reflected light. Unfortunately, this makes them usually a billion times fainter to an external observer than the star they orbit. Nevertheless, using adaptive optics, coronagraphs, and Hubble, it might be possible to do so. It has not been done yet, although some people are certainly trying. In the IR, the situation is somewhat better. The ratio of the brightness of the star to a planet is somewhat more favorable. But such a detection cannot be done just yet.

Astronometry

Another method is astrometry: looking for positional changes of a star that is tugged on by a planet. Both orbit the center of mass of the star-planet system, but the orbital motion of the star is very small. But such a “wobble” in the sky may be detectable with current technology (the Fine Guidance Sensor on the Hubble is trying to do this).

Transits/Eclipses

Yet another is to look for a drop in stellar brightness as a planet passes between the star and us. A jovian planet can cause a 1% drop.

Doppler Effect

Finally, a planet tugging on a star with give the star a velocity around the center of mass of the star-planet system. Jupiter forces our Sun to move 13 m/s.

RESULTS

So far, imaging and astrometry have not yielded any positive detections.

But, beginning a few years ago, different research groups began to detect the presence of planetary and brown dwarf companions to stars using the Doppler technique. We now know of many dozens of stars with companions, many of which seem to have Jupiter-like masses. One of these, HS 209458, has even been observed to have its planet transit the star, producing a 1.7% drop in brightness.

51 Peg, from 1 month of data in 1995	51 peg – new & improved

	Extrasolar planets detected using the Doppler technique, as of early 2007. Pretty soon this panel will be too bit to show! Fro all the details, go to http://exoplanets.org/

Hot Jupiters? Who Ordered Those?

Even a casual inspection of the chart showing the locations of detected low-mass companions reveals an apparent dichotomy. The majority of the detected jovian-mass objects are located very close to their stars, not beyond the snow line! What is going on here? Probably two things.

First, it is far easier to detect such an object if its effect on its stars is large rather than small. Because gravity is a “one-over-r-squared” force, its effects are much greater (for the same mass) if the planet is close instead of far. So, there is a bias that favors such detections.

But why should there be any such “hot jupiters” at all?

Theorists who model the formation process of planets is disks may have found a solution. If the disk material is dense enough while the planet is forming, the planet will experience “drag” as it plows thought the disk . The planet is attempting to orbit according to Kelper’s laws, while viscous friction in the disk makes the gas and dust orbit more slowly. This “head wind” that the planet experiences will cause its velocity to slow somewhat, and it will drift inward toward the star. The race between the planetary migration and the dispersal of the disk material will determine how far the planet migrates.

It may be that the conditions present in our own solar system were such that such migration was minimal. But by studying our solar system and as many others as we possibly can, we may someday get a better picture of how our planetary neighborhood came to be the way it is today.

Planets & Metallicity

The greater the abundance of heavy elements, the more likely the star will have planets. This makes sense, as it takes more than H and He to make planets as we understand it. Jovian planets contain heavy elements that may have formed a core from which the rest of the planet accreted. Certainly water ice requires lots of oxygen, and the mass of ice is probably crucial for the further accumulation of mass.
HD 209458b

Some of the hot jupiters ought to transit across the face of their star, casing a dip in the light curve. In 2002 it was announced by 2 groups that HD 209458 had such a transiting planet.

Early ground-based measurements of HD 209458. The 1.7% drop in light signaled the transit of the planet HD 209458b across the face of its star.	Properly phased light curve from HST

ANIMATIONS:

12.4 MB mpeg http://www.physics.uc.edu/~sitko/AdvancedAstro/30-ExtrasolarPlanets/transit_evap_obs.mpg

1 MB animated gif http://www.physics.uc.edu/~sitko/AdvancedAstro/30-ExtrasolarPlanets/anim_transit_quick.gif

	Despite the “boil-off”, the planet is massive enough to last billions of years.

Then came detections of gases in the atmosphere of the plane. The first was the detection of the Na I D line variability, but being only 0.02%, it was very weak. Using HST it was possible to see planet gas absorption by H I Lyman a at a whopping 15% (!!!!) from a highly extended outflow from the planet.

Wonderful “little” animation of the La “eclipse”. Maybe this is a better picture of what is going on:

http://www.physics.uc.edu/~sitko/AdvancedAstro/30-ExtrasolarPlanets/evap_e.gif

Search Strategies of the Future

Kepler http://planetquest.jpl.nasa.gov/Kepler/kepler_index.cfm	Space Interfeometry Mission (SIM) http://planetquest.jpl.nasa.gov/SIM/sim_index.cfm

Darwin http://sci.esa.int/science-e/www/area/index.cfm?fareaid=28

	Simulated interferometric image of our own solar system, The three bright spots would be Mercury, Nenus, and Earth. The Sun is removed by interferometric nulling.

Terrestrial Planet Finder (TPF) http://planetquest.jpl.nasa.gov/TPF/tpf_index.cfm

IR coronagraph concept	Fixed interferometer concept

Free-flyer concept (like Darwin)

You can find animations of the free-flyer concept in either QuickTime or Real Player formats here.

http://planetquest.jpl.nasa.gov/gallery/gallery_index.cfm

New technique – use a combination circular “bar-code” aperture mask, coronagraph, then roll to 480 different orientations to reduce effects of detector noise.

Simulation: The three objects here are intended to represent planets as bright as Jupiter, half as bright as Jupiter, and as bright as the Earth,