Tag Archives: red giants

The Double Red Giant With Odd Oscillations

My latest paper was accepted for publication on December 31, announced on the astro-ph preprint arXiv on January 5, and will appear soon in The Astrophysical Journal [UPDATE Feb 11! Rawls et al. 2016, ApJ, 818, 108]. It was written online using Authorea. I also wrote several programs in python to analyze data for this paper.

Picture two nearly-identical stars orbiting each other. Something like this:

An approximation of the double red giant binary KIC 9246715, created using an eclipsing binary simulator. In reality, the two stars are separated by more than 200 times our Sun’s radius, but this simulator maxes out at 60. One orbit takes 170 days rather than the 26 illustrated here. The brightness, or flux, dips when one star passes in front of the other. The eclipses are not evenly spaced in time because the orbit is eccentric.

Even though they usually appear as a single dot of light, binary stars are one of the best tools astronomers have to measure stellar properties. Thanks to the math behind gravity, we can weigh pairs of stars using the relationship between how long an orbit takes and how far apart the things are doing the orbiting. Weighing stars accurately is important because a star’s mass seals its fate. So if every star in the night sky had a secret companion star (or exoplanet!), we could wrap this up pretty neatly and move on to deeper questions about stars’ lives.

Unfortunately, only about half of stars have orbiting companions, and many of those aren’t observable because they don’t happen to be edge-on like the case shown above. This is where starquakes come in.

Weighing stars from the inside out

Some stars, including our own Sun, ring like bells. Pressure waves are excited by convection inside stars, and the waves bounce around at resonant frequencies just enough to make them pulse, or oscillate. Because an oscillating star is changing brightness ever-so-slightly, we can use regular observations of brightness versus time (from Kepler, in my case) to pull out the frequencies of oscillation. Heavy stars oscillate differently from lighter ones, and big stars oscillate differently from smaller ones. Voila—a new technique for weighing stars that doesn’t require anything in orbit!

Of course, the story doesn’t end there. While the study of starquakes (more formally known as asteroseismology) is a powerful way to characterize many stars quickly, it remains relatively untested. We don’t know how accurate of a scale we’re using when we whip out asteroseismology to weigh stars. To address this, my colleagues and I identified about twenty binary systems containing red giants. That’s the kind of star our Sun will become when it runs out of fuel in billions of years. Red giants are convenient targets for asteroseismology because they are bright and oscillate slowly. Both properties make them easier to observe than Sun-like stars. And since our red giant stars all live in binaries, we should be able to weigh them in two independent ways and compare the results.

Solar-like oscillations make a comb-shaped pattern at different resonant frequencies. The central location of the spikes and their spacing in frequency tell us about a star’s average surface gravity and density, respectively. Red is oscillations from the double red giant binary in my paper. Gray is oscillations of a single red giant star with similar properties, plotted upside-down for reference. Figure 7 from Rawls et al.

The case of the missing oscillations

In my paper, I present a case study of two red giant stars in an eclipsing binary. From binary modeling, I show that the stars are both a little more than two times as massive as the Sun, and over eight times as large. However, I am surprised to only find a single signature of starquakes in the observations. Two similar but not-quite-identical stars should, in principle, both oscillate. The oscillation modes, pictured above, are broader and weaker than expected, too. The same physical process could be fully stopping oscillations in one star and only partially suppressing them in the other.

By harnessing many observations (both images and spectra) and modeling techniques, I thoroughly characterize both stars and investigate why only one of them appears to oscillate. I measure each star’s mass, size, temperature, chemical composition, level of magnetic activity, and tidal force strength, among other things. Then I bring in asteroseismology to see if I can tell which star is oscillating and if its story checks out.

By simultaneously modeling different observations of this double red giant binary, I can map the geometry of its orbit in space and measure each star’s mass and size. I do this by fitting a model (black) to observations (red and yellow). The top panel shows radial velocities, or how fast each star is moving toward/away from us, and the other panels show a light curve, or brightness versus time, with two eclipses (the bottom is a zoomed view of each eclipse). The x-axis is in units of how long one orbit takes, which is about 170 days for this binary. Figure 6 from Rawls et al.
Stars emit different amounts of light (flux) at different colors (wavelengths), but the light from a binary contains overlapping information about both stars. As the stars orbit, characteristic dips in flux shift to higher or lower wavelengths depending on how fast they are moving. To create one representative spectrum for each star as shown here (red and yellow), I used a physical model of the stars’ orbit to remove the velocity offset from each observation and then combined them. This process is called disentangling. An example of a single observation containing light from both stars (before disentangling) is plotted in black. Figure 5 from Rawls et al.

As it turns out, the two stars in this binary are similar enough that it’s impossible to say for sure which one the oscillations belong to. Recent work has shown that magnetic fields may suppress oscillations in stars, however, so I strongly suspect the oscillating star is the less magnetically active of the pair. There may be a weak second set of oscillations, but the signal is very noisy and doesn’t appear quite where it should. Either way, the single mass and radius derived from asteroseismology is consistent with that of both stars from binary modeling.

Fraternal twins: born together, but not identical

Because oscillations bounce around inside stars, they carry information about how stellar interiors are structured. Stars of different ages have very different things happening inside: younger red giants are still fueled by hydrogen, while older ones are fueled by helium. The oscillating star in this binary appears to be in an advanced helium-burning stage of its life called the horizontal branch or secondary red clump. I verify this with stellar evolution modeling, and confirm that the two stars most likely were born, grew up, and evolved together. They are about 940 million years old.

The next step is to do a similar analysis for the other red giant binaries my team identified. We are working on two fronts: comparing masses and radii from binary modeling and asteroseismology, and using those results to investigate why about a third of red giants don’t show any oscillation behavior. Our work has important implications for understanding the composition of our Milky Way galaxy, because bright red giants are often surveyed to better understand our galaxy’s history and structure. It’s important to get their stories right.

Giants of Eclipse Wrapup

Interferometry, asteroseismology, heartbeats, tomography… oh my!

The rest of the Giants of Eclipse meeting saw a much wider array of subjects than just Epsilon Aurigae. We heard about interferometry, a special technique often used by radio telescopes to get sharper, higher-resolution pictures. Daniel Huber gave a great overview of asteroseismology, or stellar pulsations, which related to my talk the next day. Andrej Prsa discussed lots of work being done by the Kepler team with eclipsing binaries. The Kepler spacecraft is one of my favorites – it spent over three years collecting extremely precise data to search for planets around other stars. (Unfortunately it recently stopped working, and prospects of getting it up and running again are slim, but there is still tons of data to pore over.) A great benefit of all the high-precision Kepler data isn’t just planet hunting – it’s also extremely useful to study stars. We heard about a special kind of binary star that shows a “heartbeat-like” pattern in its light curve, and we also learned about an innovative technique called tomography, which essentially creates a 3D map from a series of 2D slices.

Brightness versus Time for a so-called "heartbeat star." The black dots are the observations from Kepler, and each line is a different model. The brightness changes because the two stars pass very close to one another as they orbit and their shapes are briefly distorted. Figure from Beck et al. 2013.
Brightness versus Time for a so-called “heartbeat star.” The black dots are the observations from Kepler, and each line is a different model. The brightness changes because the two stars pass very close to one another as they orbit and their shapes are briefly distorted. Figure from Beck et al. 2013.

So much science! By the time I presented my research on Thursday afternoon, everyone was probably tired of listening to talk after talk after talk. But I was pleased that I didn’t go over my time, got a couple of chuckles, and was able to answer questions intelligently. Next time I give a talk at a conference, I hope I don’t have to go last, because I spent a lot of time worrying about my own presentation instead of listening intently to others’.

Overall, I was thrilled to meet so many other people who care about binary stars at least as much as I do. We shared many great meals and fun evenings. I left Monterey with new friends and a better sense of what cool research is happening with binary stars.

Giants of Eclipse, Day 2

What happened to Day 1, you may ask? Unfortunately, a canceled flight happened. To make a long story short, AstronoMerrdiff finally maxed out her travel karma and got to spend an extra day in San Diego. It worked out okay, though – I worked on my talk for Thursday, and had a great impromptu meeting with my old advisor at San Diego State.

By arriving at the conference in Monterey some 24 hours later than intended, I missed a couple of overview talks, a session on VV Cep stars (a special kind of eclipsing binary), and a particularly interesting-sounding theory session. This apparently featured red giant atmosphere modeling, red giants in eclipsing binaries (!), and working out how triple systems (three stars orbiting one another) may have formed.

Let me back up a step…

This conference is all about big stars that orbit other stars so that one routinely passes in front of the other. That’s why it’s called GIANTS of ECLIPSE! It also happens to be precisely what I’m working on right now. In particular, I’m working on a project that involves red giants in eclipsing binaries. A red giant is a huge, bright red star that has run out of hydrogen fuel. Our Sun will become a red giant near the end of its life. An eclipsing binary is a pair of stars orbiting each other that pass in front of one another as viewed from Earth. As it turns out, the talk I missed yesterday about red giant eclipsing binaries was originally scheduled for Wednesday, and got moved at the last minute. Neither the speaker nor I knew this until we each arrived.

A recurring theme at the conference so far has been “this person was going to come and give a talk, but wasn’t able to.” There have been at least four cases of this, and there are fewer than 40 people at the conference to begin with. Sometimes an absent presenter is able to Skype in and/or pre-record a talk to be played back. In other cases, they ask a colleague at the meeting to give their talk for them. And in many situations, the talk is simply withdrawn. These cancellations and substitutions are what led to schedule shuffling. Why this all has to be so last-minute and offline is beyond me.

So, this brings us to today. Today’s theme was Epsilon Aurigae (usually pronounced or-EYE-jee, with “jee” sounding like “jeans”). This funny name belongs to a particularly mysterious eclipsing system. As best we can tell, Epsilon Aurigae is a cooler F-type star and a hotter B-type star orbiting each other, with a critical twist – the hotter star is embedded in a dark, dusty disk. Epsilon Aurigae is a particularly hot topic right now among stellar astronomers because it just completed an eclipse: the dark disk hiding a hot star inside just finished passing in front of the cool star. This only happens every 27 years, so lots of people were excited to observe it and begin to understand it.

Artist's impression of Epsilon Aurigae.
Artist’s impression of Epsilon Aurigae.
Image from http://en.wikipedia.org/wiki/Epsilon_Aurigae.

The best part of a small conference like this, however, is meeting people whose names you know from papers – in person. The under-40 group is rather under-represented at Giants of Eclipse, which is unsurprising given budget woes. (Grad students and postdocs are generally hit hardest when it comes to travel funding. In my case, I’m only here because I could afford to pay out-of-pocket, and that simply isn’t possible for most early-career astronomers.) So: the few of us young folks automatically meet and shake our heads as some of the more senior astronomers use overhead projector sheets (yes, really).

Thankfully, it isn’t all slide rules – there is lots of great science being done. I’m looking forward to the next couple of days, where the talks will venture further afield than just Epsilon Aurigae. Giants and binaries and eclipses, oh my!