What one research team calls "the most famous recurrent nova in the galaxy" might whittle itself to nothing in as little as 100,000 years.

T Pyxidis is a pairing of two stars in a dysfunctional relationship. They orbit each other so closely that one of them, a white dwarf, is able to suck matter off of its companion, a red dwarf. About every 20 years — six times since 1890 — the pressure of the stolen mass builds up until the white dwarf suffers a thermonuclear meltdown and erupts in a nova, spewing the accumulated matter off into space and briefly becoming some 2,500 times brighter.

RS Ophiuchi, a recurrent nova

In this depiction of the recurrent nova RS Ophiuchi, its stars are close enough that the white dwarf (at left) can siphon mass off of its red giant companion. T Pyxidis differs from this system in that its companion is a red dwarf, not a red giant.

David A. Hardy / STFC

Most astronomers believe that white dwarfs in such binary systems continue to prey off their gravitationally weaker companions until they fatten up to 1.4 solar masses, at which point they explode as supernovae.

But recent observations proffer a darker fate for T Pyx — instead of bulking up, its white dwarf appears to be losing mass. "The most famous recurrent nova in the galaxy…is dying," conclude Joseph Patterson (Columbia University) and 13 amateur co-authors in a new study. Their evidence suggests a "suicide pact" that, in about 100,000 years, will leave the white dwarf whittled itself away to nothing or its red dwarf companion consumed entirely.

The surprising result comes from a vast archive of T Pyx's brightness measurements made over 1½ decades by the globe-spanning Center for Backyard Astrophysics network of amateur observers, which Patterson coordinates. These observations stretch back to 1998, when the CBA team determined the system's orbital period by monitoring telltale fluctuations in brightness as the two stars circled one another.

Orbital period of T Pyxidis

Observers in the global Center for Backyard Astrophysics Network have been measuring the orbital period of the binary system T Pyxidis since the late 1990s. They found that the period increased by about 0.005% (red symbol at upper right) after an outburst in 2011, signifying that the system lost mass during that event.

J. Patterson / CBA

Timing is Everything

T Pyx's most recent outburst, in 2011, gave the team a perfect opportunity to go one step further. In theory, if mass is lost, the gravitational force between the two stars is weakened, and they'll take longer to orbit each other. So by measuring the orbital period before and after the outburst, Patterson and his observers could determine how much mass had been lost.

The CBA team found that the period increased from 1.829507 hours to 1.829606 hours, an increase of 0.0054%. Although this may seem tiny, Patterson was still amazed: "So large a period change is very, very surprising," he writes — seven times larger than predicted. This corresponds to a mass loss of 1026 kg, about a Neptune's worth — five times greater than what the white dwarf gained in the decades before its outburst. "It seems unlikely," the article states, "that the white dwarf…will ever increase its mass at all, much less reach 1.4 solar masses."

Ejected shells around T Pyxidis

Left: Shells of ejected matter surround the recurrent nova T Pyxidis, about 6,000 light-years distant. This 1995 image is from ESO"s New Technology Telescope in Chile. Right: A compilation of Hubble views reveals that the shells actually consist of more than 2,000 gaseous blobs packed into an area 1 light-year across.

M. Shara & B. Williams (STScI) / R. Gilmozzi (ESO) / NASA

Patterson speculates that T Pyx is so weird due to the vulnerable nature of its companion, the lowly red dwarf. In most known recurrent novae, the mass donor is a red giant, rather than a dwarf. "The outburst does not tremendously affect a nearby red giant," explains Patterson, "but it tremendously affects a low-mass red dwarf." The red dwarf is overwhelmed, its outer layers becoming bloated by the intense radiation. The blast then drives these wispy outer layers away, instead of replenishing the white dwarf. This might explain the enormous mass loss seen during the 2011 outburst.

However, not everyone believes T Pyx is doomed to simply fade away. "It just absolutely flies in the face of theoretical expectation," says Edward Sion, an astronomer at Villanova University who is no stranger to T Pyx and its peculiarities. In 2010, he led a study that concluded the system's white dwarf is steadily gaining mass and might be only 10 million years away from going supernova. "I'm not saying Patterson's group is wrong," he said during a phone interview, "but we have to be very cautious."

Sion warns that Patterson's calculated accretion rate falls into a grey area filled with theoretical contradictions. For example, Patterson assumes that T Pyx's white dwarf has 70% of the Sun's mass. But in that case it shouldn't be accreting matter fast enough to trigger outbursts every 22 years. And yet, Sion continues, Patterson's assumed accretion rate is too high — material is piling up on the surface of the white dwarf rapidly enough to be burning constantly, not waiting decades before erupting.

Debate Advances Science

Sion has in mind a different picture of T Pyx. In his view, the white dwarf has a strong magnetic field that clears the inner regions of the surrounding accretion disk and funnels this material to its magnetic poles, creating hot spots visible in ultraviolet light. He's currently analyzing spectra of T Pyx from the Hubble Space Telescope taken late last year, and he believes the white dwarf will turn out to be much more massive, closer to the critical 1.4 solar-mass limit. If true, this model would align much more closely with theory, which predicts that such novae have smaller outbursts and larger accretion rates than what Patterson's team has found.

That would be just fine with Patterson. When predictions prove to be wrong, he comments, "they stimulate people to make measurements — and everyone wins." No matter the outcome, CBA observers advanced the scientific debate by measuring T Pyx's mass loss. As Patterson notes, "It was lovely to see an important result coming directly from good ol' Kepler's laws applied to a huge swath of data from backyard telescopes."


Mark Zastrow, a graduate student in the astronomy program at Boston University, is also a science writer and photographer. Follow him (@MarkZastrow) on Twitter.

Comments


Image of Peter

Peter

March 25, 2013 at 7:48 pm

If a longer period means less mass, then according to the chart, it was losing mass before the outburst?

You must be logged in to post a comment.

Image of Mark Zastrow

Mark Zastrow

March 25, 2013 at 10:20 pm

Hi Peter,
That's a great point. Yes, the period is steadily increasing during the quiescent phase as well, when the white dwarf is gaining mass from the red dwarf.

It may seem odd that the white dwarf can both gain and lose mass during a period increase. But the difference is that in the quiescent phase, the total mass of the system doesn't change—the matter is being transferred from the red dwarf to the white dwarf. In the outburst, the white dwarf is ejecting mass from the system entirely. In both scenarios, in order to conserve angular momentum, the result is that the stars spiral outward and the period increases.

You must be logged in to post a comment.

Image of Bruce

Bruce

March 26, 2013 at 6:22 am

Nice report Mark. I’m intrigued by the kinematics of this system and wonder if, with their close proximity, the force of the blasts during the nova outbursts may help push the pair apart. If this is the case then it wouldn’t be just mass loss that’s lifting the red dwarf into a higher orbit.

You must be logged in to post a comment.

Image of Bruce

Bruce

March 26, 2013 at 6:31 am

I that should have been kinetics, not kinematics.

You must be logged in to post a comment.

Image of Mark Zastrow

Mark Zastrow

March 26, 2013 at 8:57 am

Thanks Bruce. That's an excellent question—that was, in fact, another one of the caveats that Ed Sion brought up in our conversation that he says hasn't been extensively studied yet in the literature.

You must be logged in to post a comment.

Image of Alan Seeger

Alan Seeger

March 27, 2013 at 10:06 am

So when the outbursts occur, the mass ejected from the system must be blown out at a sufficient velocity to achieve escape velocity for the system...? It seems as though the gravity of one or the other of the companions would be sufficient to capture the ejecta into a disk of orbiting material. The orbital velocity of Earth is ~7 miles per second, so I would imagine that of this system would have to exceed that... it's pretty impressive to me that this is even possible.

You must be logged in to post a comment.

Image of Bruce

Bruce

March 27, 2013 at 9:08 pm

Alan, if you’re impressed by a mere 7 km/s then prepare to be blown away. The formula for escape velocity is Ve = sqrt (2GM/R). The mass of the white dwarf component of T Pyx is debated, but a typical white dwarf packs a sun’s worth of mass into an earth’s worth of space. Plugging in the numbers yields a Ve of around 6,450 km/s for matter at the surface of your average white dwarf, so the nova blast must leave the white dwarf at around a thousand times the earth’s orbital velocity, or maybe even more. Remember that what causes the nova is a nuclear blast across the white dwarf’s entire surface. The effect upon the red dwarf must be enormous too.

You must be logged in to post a comment.

Image of Peter

Peter

March 28, 2013 at 1:14 pm

Okay, the red dwarf is spiraling away from the white dwarf. We can assume that a long time ago, they were in a tight orbit?

You must be logged in to post a comment.

Image of Bruce

Bruce

March 28, 2013 at 11:03 pm

Aren't they already in a tight orbit Peter? I mean, a little over 109 minutes is pretty short as orbital periods go. I tried a back of the envelope calculation using Newton's form of Kepler's third law, P^2=4•pi^2•a^3/G(m1+m2) to attempt to estimate "a", the average distance between the two stars centers. Not knowing the stars' masses messed up the attempt however. I guessed the masses at 1 sun and .2 suns, but I came up with an answer that put the white dwarf INSIDE the red dwarf! Does anyone have info on what the masses are thought to be???

You must be logged in to post a comment.

Image of Peter

Peter

March 29, 2013 at 10:34 am

Yes, Bruce. I don’t think the white dwarf is orbiting inside the red dwarf, but my point is that run backwards, it suggests something like that.

You must be logged in to post a comment.

Image of Mike W. Herberich

Mike W. Herberich

March 29, 2013 at 11:02 am

Any suggestions why the increase in orbital period (refer to chart) is not linear but rather should have bumps of 4 years in duration, approximately? What could explain material to accrete (or, actually, the period to decelerate) in a way so regularly irregular between two novae? The measuring ranges indicated do not seem to lend themselves to explain this.

You must be logged in to post a comment.

Image of Bruce

Bruce

March 29, 2013 at 5:57 pm

No, of course the WD isn’t inside the RD Peter. And I found the error I was making in applying Kepler’s third. For my test example of a 1 sun mass WD & a .2 sun mass RD the average separation with this system’s short period is around 560,000 km, leaving > half a million km between the two stars. But yes Peter there very like was at least one period when this system was a contact binary. Or maybe there was a time in their history when the present WD was a red giant and the RD spiraled into a close orbit? Speculating further, maybe the present RD was a gas giant or a brown dwarf at one time but received an upgrade as Red Giant gas was dumped onto it. Mike H., welcome back to the conversations. Could what you’ve noticed be caused by oscillations in the rate of mass outflow from the RD? This system is, as Spock used to say, “Fascinating.”

You must be logged in to post a comment.

Image of Peter

Peter

March 30, 2013 at 8:29 am

Thanks, Bruce, but I am really looking for an answer from the PhDs. Yes, one can easily imagine they were once a contact binary, but one cannot imagine them separating from that point. There is no known force of nature that could separate a contact binary, once formed. One cannot argue with the observations of an increasing period, but the model points to a physical impossibility in the not-too-distant past. Either something is fundamentally flawed with the model, or there is some new anti-gravity force of nature waiting to be discovered. Fascinating, indeed. Perhaps there is an unusual concentration of dark energy in T Pyx's vicinity?

You must be logged in to post a comment.

Image of Bruce D. Mayfield, non-PhD

Bruce D. Mayfield, non-PhD

March 31, 2013 at 8:44 am

Well Peter, please excuse me for not having any letters after my name, but just as amateur astronomers helped provide the data that may decode this mystery, this amateur armchair astronomer would like to offer up an idea of what may have happened in T Pyxidis’ past. (The thought occurs that someone seeking a doctorial thesis in astrophysics might consider solving the riddle of this system.) Consider the possibility that T Pyx starts out as an ordinary main sequence primary (A) orbited at some critical distance by a RD (or less) massed secondary (B). T Pyx A burns through its core H and swells to the point that it fills its Roche lobe and starts dumping its outer envelope onto B. Since B’s mass is considerably < A’s the Roche lobe on B’s side will fill as well and the pair becomes a contact binary. “A” swells further and the pair may briefly share a common envelope, causing B to spiral in somewhat, but before the cores combine A’s core exhausts its fuel and its core contracts into a WD while it blows off its outer envelope. Now the system is a WD being tightly orbited by a RD evolving as we see today. This picture is undoubtedly simpler than what actually occurred here, but I think the basic outline is at least plausible. No strange physics is required to decouple a contact binary Peter. All that’s required is for the primary to run out of fuel before the cores combine.

You must be logged in to post a comment.

Image of Peter

Peter

March 31, 2013 at 3:00 pm

Okay. I made a fool of myself with that last post. Your explanation makes perfect sense.

You must be logged in to post a comment.

Image of Bruce

Bruce

March 31, 2013 at 7:17 pm

Been there, done that too Peter. You're no fool. It's not foolish to suggest something outside the box as long as you retain the ability to let go of it if its shown to be untenable.

You must be logged in to post a comment.

You must be logged in to post a comment.