20 Ring Worlds

image credit: Credit: ALMA (ESO/NAOJ/NRAO), S. Andrews et al. In a team of 17 scientists lead by Sean Andrews, we carried out the Disk Substructures at High Spatial Resolution Project, or DSHARP in short. This consists of over 60 hours of ALMA observations for 20 planet forming disks designed to study the occurrence, forms, locations, sizes, and other properties of structures in the disk material and how they might be related to the planet formation process.

We found that rings are everywhere, while spirals and asymmetries are sometimes present but rare. The positions of the rings show no correlation with the stellar brightness, which basically rules out that condensation fronts are responsible for the rings. Together with the other recent ALMA result (see three baby planets detected) this seems to indicate the planets form abundantly and faster than expected. At the same time, these results also open new questions, for example: why do all rings seem to be equally transparent? When do these structures (or planets) form, are they already present in much younger disks? What to disks in other star forming regions, that are less massive look like? We still have so much to learn ... ćA; You can read more about this

Three baby planets detected

image credit: Credit: NRAO/AUI/NSF; S. Dagnello In Rich Teagues paper on HD163296 we used the ALMA observatory to hunt for baby planets in their natal environment, the planet forming disks. We use the Doppler effect to measure the gas rotation velocity with unprecedented precision of only a few meters per second. Keep in mind that the disk is about 300 light years away and rotates at speeds of a few kilometers per second! Planets forming inside those disks perturb the gas and cause it to rotate slightly faster or slower. The detected rotation curves are a perfect match to predictions from planet-disk interaction, as coauthor Jaehan Bae carries out. This is good evidence that indeed planets of about Jupiter mass have already formed in this particular disk around the young star HD163296. This is quite surprising, as astronomers have no idea how these massive planets formed so far away from their star (they are more than 80 times farther away from their star than the earth is from the sun). Next, we will apply this method to other disks to see if all the dust rings we see are caused by planets. You can read more about this

Press Releases on TW Hya and HL Tau

image credit: Carrasco-Gonzalez et al.; Bill Saxton, NRAO/AUI/NSF The last two weeks were busy - we released two press releases, both happened to be about observed substructure in disks.
The first one is on the iconic HL Tauri disk that we observed with the EVLA at optically thin wavelengths where we resolved a possible clump in the innermost ring.
For more information, you can read The second press release is about our ALMA observations of the TW Hya disk. These are pretty much the highest resolution observations ALMA can do aimed at the closest disk to us. Observations in the foreseeable future will not get much better than this. We found surprising ringed substructure, that is several ring/gap features, much narrower than seen in HL Tau. In addition to that, the inner gap that was predicted from analysis of the spectral intensity distribution before was finally imaged and found to be of 1 AU size. It is exciting to consider that these are likely the signposts of planets forming in the innermost regions of the system.

You can check out image credit: image credit: S. Andrews (Harvard-Smithsonian CfA), ALMA (ESO/NAOJ/NRAO)

Colder than expected

image credit: Digitized Sky Survey 2/NASA/ESA, inset: ESO/NASA/ESA The image to the left shows a circumstellar disk that is nick named "the flying saucer". The star in the center of the disk is not seen directly, because the thick dusty disk is exactly seen from the side (it is the dark line in the image) and it absorbs all the light that the star shines directly at us. The bright regions above and below that dark line is light that scatters off small dust grains above and below the disk. In addition to being so exactly edge-on, there is another peculiar arrangement: the disk is in front of a cloud of dust and gas. This special arrangement allowed us to directly measure the temperature of the dust using the ALMA telescope array.
It is commonly believed that dust in circumstellar disks, even far away from its central star, cannot be colder than about 15 degrees above absolute zero because there is always interstellar radiation that heats the dust particles to this minimum temperature. With this in mind, our measurements came as a surprise: we measured a dust temperature of only 7 degrees above absolute zero! There are some possible ways to explain this, but they all indicate that the dust grains have different shapes or different optical properties from what was commonly assumed.
Why does that matter? If we want to understand where planets can form and how large they can become, we need to know how much material is out there, in other words, we need to weight the disk. The methods to measure the masses of disks however rely on the knowledge of the dust temperature and other properties of the dust. Our measurements show that the dust seems to be quite different from what astronomers assumed it to be so far and they highlight why understanding the physics of those small dust particles is such an important part of understanding the origins of planets.

MPIA press release: English | German
ESO press release: English | German

Pre-transition Disks

image credit: P. Pinilla The origin of pre-transition disks, disks around young stars that consist of an outer and an inner dust ring, have puzzled researchers for decades. The inner rings seem to live very long, even though scientists did not think this could be possible. Our recent paper, published in Astronomy & Astrophysics, proposes a solution to this puzzle: the outer ring constantly supplies the inner ring, but it does so without being seen: the material is transported in only few large grains that are not easily detected. Therefore the regions between the rings do appear to be empty. Once the dust particles have reached the hot inner regions close to the star, they loose the water ice that binds them together and consequently, they break up into many small pieces. It is these small pieces we see close to the star as a second, inner hot ring of dust. Our paper was featured on (Dutch).

National Geographic

National Geographic featured an article called Cosmic Dawn which tells the story of the ALMA Observatory and how it opens up new views into the early universe and the formation of stars and planets. It also mentions our Science paper on the "dust trap" in IRS 48. The article is beautifully designed and I highly recommend to check it out!

Recent papers, Episode 5

image credit: ALMA/Calcada

It's a trap!

Recently, I was involved in two related papers:

Lopsided dust rings in transition disks
T. Birnstiel, C. P. Dullemond, P. Pinilla, A&A (2013) vol. 550, L8

A Major Asymmetric Dust Trap in a Transition Disk
N. van der Marel et al. Science (2013) vol. 340, 1199-1202

In the first one, we proposed that the rings seen in transition disks can easily become asymmetric: only a small asymmetry of the gas bump is enough to trigger a strong accumulation of dust particles.

Right around the time we were finishing up this paper, we were contacted by some colleagues who had just gotten their latest data from the ALMA observatory. They wanted to investigate the gas structure of the disk, but then they found a much more interesting and surprising feature in the dust which they had not expected: instead of a symmetric dust ring, they saw an extremely asymmetric structure, just like we predicted. We joined them in interpreting the data and the paper was published in the journal Science.

You can read more about it in the press releases from NRAO or ESO, or in German from the University of Heidelberg. It was also covered in an ESOCast - the video can be found here.

Recent papers, Episode 4

Forming the building blocks of planets by winning the lottery

A summary of the following nice paper by Fredrik was long overdue:

Breaking through: The effects of a velocity distribution on barriers to dust growth
F. Windmark, T. Birnstiel, C. W. Ormel, and C. P. Dullemond, A&A (2012) vol. 544, L16

In these two papers, we show how the building blocks of planets can be formed by mutual sticking instead of gravitational attraction, in contrast to recent beliefs. The short explanation is that we include two key effects, one is based on recent laboratory results (see also this post), the other one involves randomness, such that a few "lucky" particles can continue to grow, even in a generally hostile environment.

However with this review, I'm really lazy and instead of explaining the details here, I let the first author speak for himself, so head over to Fredriks blog, to read how we solved these problems.

Recent papers, Episode 3

Double Feature on Transition Disks:

This one covers two papers on a similar topic:

Ring shaped dust accumulation in transition disks
P. Pinilla, M. Bensity, and T. Birnstiel, A&A (2012) vol. 545, A81

Can grain growth explain transition disks?
T. Birnstiel, S. Andrews, and B. Ercolano, A&A (2012) vol. 544, A79

So the obvious topic is "transition disks". Transition disk really means "disk with a large inner hole". About 99% of the mass of the disk is molecular (H2) gas which is mostly invisible and only 1% is dust. But it is the dust which is mostly observed and where inner cavities are found (wether the gas is depleted or not is not yet clear). These holes are detected either through imaging in the millimeter wavelength range or by looking at how much energy is emitted at different wavelength ranges. The latter one is called the SED (spectral energy distribution). This is what it schematically looks like:

Sketch of an SED

The big yellow spectrum is the one of the star which is dominating the emission, while the disk radiates at different, lower temperatures. At wavelengths longer than a few micrometers, the emission of the disk becomes dominant, this is the infrared excess. The total sum of the star+disk spectrum then composes the SED (black solid line).

Now what happens if you take away some disk material close to the star? Imagine you remove the orange spectrum which corresponds to the warm dust inside of couple of AU: then you get the dashed black line instead, so there is a dip in the SED. That is what people have looked into during the last two decades. But since a few years, there are powerful millimeter wave interferometers (e.g., the SMA and soon ALMA). These instruments have the sensitivity and angular resolution (and enough baselines) to make images at radio wavelengths, which show us how the dust is distributed in the disk. And indeed the dust clearing was also found in the images.

So what is the physics behind these signatures? The theory of accretion disks predicts a slow fading away of the disk, which is not what is observed. Some of the first suggestions were that the radiation from the star starts to evaporate the disk from the inside out. This does indeed work, but it turns out that some of these disks have holes which are much larger than what could be explained this way, and there are a few other problems with this idea. But thankfully, there are two more explanations floating around: these effects could be caused by the fact that dust grains grow and start to become invisible or what we see is the signature carved by a forming planet.

In these two papers, we investigated those two ideas. Concerning the effects of growing grains alone, we found that they can indeed produce an SED which would look like the one of a transition disk, however the images don't look like the observed mm images because the grains cannot grow large enough to be invisible to the mm observations.

However, the idea of a companion seems to work: a massive planet embedded in the disk gravitationally perturbs the disk which leads to a density bump outside of the orbit of the planet. As discussed in an earlier post, such a pressure bump can trap large particles which would otherwise spiral inward towards the star. The bump thus prevents them from reaching the inner disk, while the grains which are present in the inner disk continue to spiral inward - a hole is formed. We therefore think that a companion (planet or brown dwarf) is the most likely explanation of these observations - let's see what future observations will reveal!

Recent papers, Episode 2

Still catching up with some previous papers, this time it is Paolas first paper:

Trapping dust particles in the outer regions of protoplanetary disks
P. Pinilla et al., A&A (2012) vol. 538, A114

Dust particles in circumstellar disks are expected to collide and stick to each other, thus growing from sub-micrometer in size to planets. However this is not as easy as it sounds because apart from growth barriers (see last post), there is also an effect, called radial drift. Once particles have reached a certain size, they start to decouple from the gas flow and as a consequence they spiral inward. The size at which that happens reaches from roughly one meter in the inner disk (for example at the distance of the Earth) to particles of only millimeters or less in the outer disk (say around 100 times the Earth-Sun distance).

Now this effect of radial drift is quite simple physics, so we would be quite certain that this should indeed be at work in disks (possibly slightly weaker or stronger than we might expect), but the real problem comes from observations: observatories like the SMA, CARMA, or also the upcoming ALMA are used to detect and characterize these disks in the (sub-)millimeter wavelength range, which is sensitive to dust emission, particularly to grains of around millimeters in size. Several people have found that particularly the outer disk is full of mm or cm sized particles, exactly those which shouldn't be there according to the expectations of radial drift, so something has to halt or suppress radial drift.

The most straightforward way to stop radial drift is changing the pressure gradient in the disk: the closer you get to the star, the denser and hotter the gas becomes, so pressure is increasing as you get closer to the central star. Now the drift speed scales linear with the pressure gradient, which is the rate at which the pressure de- or increases. Therefore, if you have a region in the disk, where the pressure is constant, there is no drift. Taking this one step further, if there is a region where pressure is increasing with distance to the star, then particles should drift outwards instead of inwards.

But how do you do that? Now there are a few ideas out there ranging from turbulent over-densities, or spiral arms to more complicated effects such as zonal flows. All these effects are disturbances to the density structure of the disk, so our idea with this paper was to parameterize the disturbances and test what size and strength of the perturbation is needed to influence radial drift and growth of dust particles and how observable quantities are influenced.

What we found was the following: the best option to have efficient trapping of dust particles are sizes of around one pressure scale height (this is just a typical length scale for disk physics) and an over density of at least 30%. With values like these, exactly those grains are kept in the disk which are needed to explain observations. The plot shows the spectral index, which is a proxy for grain size (lower alpha means larger grains) versus the total flux, which is a measure of dust mass (more flux means more dust). The dots represent the observational data, the red area is what our theoretical models can cover if there is no drift at all. If radial drift were active, then our models would predict larger alpha and less flux as time proceeds - this obviously goes in the wrong direction, as you can see by looking at the orange arrow. Our new models with particle trapping (symbolized by the green arrow) manage to get both quantities right: the dust mass decreases slowly with time (since the flux goes down) while at the same time, particles are allowed to grow to larger sizes and keeping this size over time.

spectral slope vs millimeter flux

This paper was also selected as Highlight Paper by Astronomy & Astrophysics. Check out the A&A website for their shorter summary.

Recent papers, Episode 1

From now on, each time something gets published with my name on it, I will try to post some (hopefully easy to understand) explanation about what we have been doing. So here's my first post about Fredriks recent paper:

Planetesimal formation by sweep-up: How the bouncing barrier can be beneficial to growth
F. Windmark et al. A&A (2012) vol. 540, A148

Dust grains in protoplanetary disks grow due to sticking collisions and at some point are supposed to become the precursors of planets. However there are a couple of issues: growth barriers and radial drift. Ignoring the drift issue one for now, particles grow until at some specific particle size, they do not stick upon colliding, but instead they just bounce off or fragment each other.

In this work, we investigated how recent laboratory work on dust collision physics can be implemented in our numerical models to possibly circumvent at least some of these problems. Laboratory work has shown that high-velocity impacts of small particles can cause two effects: erosion (like sandblasting) or mass-transfer fragmentation (i.e. the impactor is shattered but still deposits some mass on the target). So if there is enough small dust hitting the larger particles at the right velocity (so that they add mass instead of eroding the target), the target could continue to grow.

Fredriks results have shown that with the current collision model, particles still get stuck at small sizes, just bouncing off each other. These particles are however of the right size to be swept up by larger ones. So if, for some reason, particles of ~ 1 cm in size were present, they would continue to grow through this sweep up process. Only a few "seeds" are needed because once these seeds have grown to larger bodies, they would produce fragments which are also able to grow through the same process. Bodies of around 100 m in size could be formed by this process, the origin of the seeds, however, is yet to be explained.