Tuesday, October 07, 2014 ... Deutsch/Español/Related posts from blogosphere

Hubeny, Semenoff: eternally accelerating quark breeds its antiquark

Off-topic, Nobel prize: the 2014 physics Nobel prize goes to three Japanese guys for the discovery of blue LED diodes. Nakamura with or without Tokura were guessed in 2009 and 2005 so it's no "total surprise" but I still think that there have been way too many prizes like that. Blue LEDs are undoubtedly "electronic engineering" – that's where Nakamura also got his degree – and not really physics and because the advance was really about combining three chemical elements, a chemistry prize would be more appropriate.
Veronika Hubený (the surname means male-and-skinny in Czech; of course that her name would be female-and-skinny, i.e. Hubená, if Veronika's family emigrated after her name was settled LOL) and Gordon Semenoff published two playful and potentially important hep-th papers today,
Holographic accelerated heavy quark-anti-quark pair (4 pages)

String worldsheet for accelerating quark (28 pages).
The long paper boasts some really pretty colorful 3D images. But I want to introduce you to their claims most quickly so let me post this simple diagram instead:

The red quark uniformly accelerates along a hyperbola in the Minkowski space. Its trajectory looks like the straight vertical line. The world line isn't a geodesic but you may imagine this trajectory to be allowed in the real world, e.g. in the presence of a uniform electric field. Normally allowed trajectories begin in the bottom corner of the Penrose diagram; but here we can start from the "edge" (\(\mathscr{I}^-\) or "scri-minus") because of the acceleration.

Now, the quark has a one-dimensional world line. However, they think about a similar situation in the AdS space where this world line is a boundary of a two-dimensional world sheet. The "added dimension" should be identified with the single dimension that holography always adds. The "right world sheet" is determined as a "minimal surface" and they study its overall topology. Their methodology is therefore some conceptually simple geometry of surfaces in the Minkowski or AdS space.

The surprising conclusion, although a bit debatable one, is that if the red quark trajectory has been eternally accelerating in the past, the world sheet would be "incomplete" by itself and it must be continued in such a way than a green antiquark on the other side of the Universe had to be there "before" the red quark was born in the infinite past!

So they say that in some sense, the green world line of the antiquark is an unavoidable companion of the accelerating red quark! It may remind you of some comments by Wheeler and Feynman that not only antiparticles are particles moving backwards in time; maybe, all particles and antiparticles from the same species may be connected into a single trajectory going back and forth in the spacetime. ;-) Hubeny and Semenoff tell you that they could also be connected through \(\mathscr{I}^\pm\), too.

I do believe that "something like that" has to be true (after all, various correlations between the opposite sides of a Rindler wedge have appeared at many places) but I am not 100% sure whether they know completely accurately what the right statement of this kind is. The green antiquark lives behind the Rindler horizon of the red quark. I would add and claim that it is needed for consistency, to avoid contradictions from closed time-like curves. The point is that the green antiquark world line may be interpreted as the "past" before the red quark came to existence at \(t=-\infty\). Yes, there can be something "before the infinite past". The situation may look symmetric and the green antiquark may also arise "after the life of the red quark" at \(t=+\infty\). However, it doesn't have to.

Meanwhile, if you only consider the acceleration of the red quark to be past-eternal, the red quark can't influence any event on the green antiquark world line because those are spacelike-separated. I suppose that the opposite claim to theirs must also "separately" hold. If you guarantee that the red quark accelerates eternally in the future, it will get reincarnated as an antiquark in the past across the Rindler horizon. ;-)

The papers have many more claims and the short paper evaluates some quark-antiquark entanglement entropy to be \(\sqrt{\lambda}\), too.

I surely do think that the people studying the black hole information puzzle should spend at least hours by trying to think about these claims as carefully as they can. Ideas directly analogous to these Hubeny-Semenoff claims could have applications for spacetimes with black hole horizons. It could tell us that in some sense, the matter that collapsed to the black hole could be viewed as a part of the "future of a singularity" or something like that. Alternatively, the Hawking radiation could be thought of as "past-eternally accelerating matter" so it could be a continuation of some seemingly disconnected antimatter that ended in the singularity, too.

These are potentially confusing ideas analogous to but different from the "black hole final state" and people should exploit the chance to find a claim of this sort that is more correct than the previous ones. I strongly believe that similar qualitatively surprising insights about different topologies and connections between manifolds and objects in spacetimes with horizons will ultimately profoundly deepen our understanding of quantum gravity.

Add to del.icio.us Digg this Add to reddit

snail feedback (33) :

reader Ben said...

Dear Lubos,
I have been following your blog for many years now, and never commented (also here I have nothing of value to add). I would just like to say that I appreciate your work a lot - many of your physics discussions are much more lucid than the average peer-reviewed hep-th article. Please keep up the good work!

P.S.: Are you still actively doing research? I understand that you don't engage with the publication procedure to communicate intermediate results.

reader Peter F. said...

I just know, intuitively, that What Is (ultimately going on) is a combination of fundamentally finite and infinite aspects. What the precise mathematical description of some or all of these aspects are, I leave to (from) skinny (to fat) and more or less male or female physicists to try figure out. ;-}

reader Luboš Motl said...

The SSC was popular in Texas but it was killed, nevertheless, largely because it had lost the key supporter in the White House which was only Reagan. Bush 41 was already calmer and Clinton-Gore were indifferent. People love to invent lots of other stories but it wasn't about the broader public and it shouldn't have been about the broader public.

reader W.A. Zajc said...

Thanks for the pointer to this very interesting paper, Lubos. I agree with your description of their "really pretty" 3D images of the world sheet; I hope I will be able to understand enough of the paper to understand if their analysis itself is really pretty. BTW, I do not mean to trivialize their wonderful figures - they certainly help me understand the nature of the solution they have found. It is also nice to see that it builds on the work of a very impressive former Columbia student Bow-en Xiao (their Reference 5): B.-W. Xiao, “On the exact solution of the accelerating string in AdS(5) space,” Phys.Lett. B665 (2008) 173–177, http://arxiv.org/abs/arXiv:0804.1343 .

reader Luboš Motl said...

Well, you say that the standard deviation is "just ignorance". But if you admit that "f" is changing for various reasons, a part of the uncertainty in "f" may be due to these changes, too, right?

I think that it is an important question how much "f" was changing over the geological history, when the atmosphere had vastly different compositions, when the Antarctica didn't cover the South pole, and so on. If those changes were capable to change "f" at least by 0.3 in both directions (or at least in the dangerous one), at least for some rare periods, than f=0.7 (today) is excluded.

reader lukelea said...

Dear Lubos, If I understand, you mentioned something about the possibility of a few hints of supersymmetry coming out of LHC after it ramps up to maximum energy next, but how how this would only add to the argument in favor of a new collider in order to flesh out the possibilities at higher energies. My question is, won't it be much better if those first hints do appear than not? I mean in terms of their PR value, which would make it easier to persuade the politicians who make the final decisions? In that admittedly unscientific sense, doesn't it make sense to root for some positive results?

reader lukelea said...

Dear Lubs, as you say, big data may be done well by the financial folks. I question however whether this has been of much, if any, real economic value for society as a whole. Some slight arbitraging advantage for the firms involved, who leverage their bets using other people's money. At least this is my impression.

reader Curious George said...

I wonder why this applies to a quark and not to an electron.

reader Luboš Motl said...

It is the N=4 gauge theory so all the particles are colored I.e. adjoint.

reader Luboš Motl said...

Of course that discoveries at any moment are good for PR and self-confidence.

reader Luboš Motl said...

Dear Luke, the big data folks are doing many other things than trading with derivatives or things like that. They quantify risk of default for loans or insurance and hundreds of other things. A big part of the modern economy if not most of it is related to these activities. These people have very high salaries and often for a good reason. Too ban these occupations would more or less mean to introduce communism i.e. to reduce the GDP by an order of magnitude.

reader OneStringToRuleThemAll said...

"Veronika Hubený (the surname means male-and-skinny in Czech; of course
that her name would be female-and-skinny, i.e. Hubená, if Veronika's
family emigrated after her name was settled LOL)" This is the main thing I got from this post, thx Lubos.

reader Luboš Motl said...

You are welcome. She told me that she was a kid and it was better to have an identical surname as her father. It makes sense. Of course that adult women who emigrate keep the feminine version of the name, like Martina Navratilova. The masculine name is Navrátil.

reader Honza said...

Hi Lubosi, I was wondering how much and in what way the fudge factor "f" depends on temperature. It definitely changes at 0 and 100 C, but how does it behave between those points?

reader Gordon said...

Hey, relax--don't be so literal---I was winging it...:)

reader Dilaton said...

Are the quark and the anti quark of the second longer paper some kind of entangled too...?

reader Jim H. said...

Oh Christ, another deep wonderful idea for cranks to abuse. ;)

reader Luboš Motl said...

Dear Dilaton, a great question.

Their analysis is classical so the spin isn't really there but I think that it's obvious that if the presence of the antiparticle is imposed, the spin has to be determined, too. Well, it's not really "entangled". If you decide that your local quark has spin-up, there won't be an entanglement because the entanglement needs both "up" and "down" to have a nonzero probability.

However, it's really correlation/entanglement "behind the horizon". In some sense, they're arguing that the physics behind the horizon is indeed determined by the degrees inside the horizon, and they do so in the most trivial way - some kind of an image. I have doubts about any such a simple solution. The horizon itself should be thought of as having a huge entropy by itself, so there should be many ways to pair the two regions, and so on.

I may be wrong but I wouldn't recommend to trust all of their interpretations literally.

Maybe, their statements become "more accurate" near the AdS boundary, i.e. in the UV, in some sense.

reader Luboš Motl said...

OK, I will relax ;-) - and when we receive the bill to pay $10 billion for the money spent for HEP on the SSC, I will send the bill to you to pay it. ;-)

reader Luboš Motl said...

Dear Honzo, an extremely good question, but an ill-defined one, too.

Note that "f" is some coefficient in a model for the global mean temperature - so it isn't a part of a mechanism that only operates locally. The feedbacks have different contributions in different regions of the world, and those regions cooperate in various feedbacks via circulations etc.

Because there is nothing such as a single global temperature, you can't talk about the dependence of "f" on "the" temperature. There is no "the" temperature. Some places of the globe are frozen, others are nearly boiling.

More properly, the whole model with a single "f" is of course bound to be inadequate whatever "f" is. The CO2-caused warming at different places etc. is described by some processes with different values of "f", and when we talk about "one f", it's again just some kind of a global average, and I don't say arithmetic average.

With these disclaimers said, I think it's obvious that exactly because there are these phase transitions and they (at least the freezing) influence *some* places of the globe, it's important how big this part is, so the effective "f" will depend on the "global mean temperature". But of course that the dependence won't be as sharp as the dependence would be if you had an "f" for one place with a single well-defined temperature. Effectively, the dependence will be smoother because different global mean temperatures effectively correspond to different percentages of the globe below or above the phase transitions etc.

At the end, the temperature changes, if very high, will influence the CO2 concentration, too. Glaciation cycles saw the CO2 between 180 and 280 ppm. During the maxima and minima, the climate sensitivity and therefore "f" was probably also different from each other.

As far as I know, the "accurate" knowledge of the sensitivity is virtually non-existent even if you talk about the single global normal-conditions estimate, so papers debating the dependence of the climate sensitivity on all these complex issues don't exist at all, not even attempts.

reader Honza said...

Thanks for the answer. If I understand it correctly, not only my question but the whole sensitivity concept is ill-defined. That is a major problem climatologists should be discussing from dawn to dusk, if not 24/7. ;-(

reader Luboš Motl said...

Dear Honzo, I wouldn't quite say that the sensitivities are ill-defined. Within some broad given conditions, it should be. The transition from one well-defined state to another - everything specified, CO2 changed from 280 to 560 ppm, for example - leads to some in principle calculable warming, up to unpredictably noise etc., if everything is equal.

I was just saying that the simple model that this warming comes from 1+f+f^2+f^3 etc. for a fixed "f", i.e. the iterative application of a well-defined feedback, isn't quite exact. So the time dependence of the warming and other things will also fail to agree with the simple one-dimensional model of the perturbation etc.

reader AHD said...

Somewhat off-topic, but I didn't see where else to ask the climate-related question: Where can I find a time-series of total sea ice volume? Yes, arctic sea ice volume is low by recent historical standards but antarctic sea ice is >2sigma above recent means. And their changes are anti-correlated so the evolution of their sum should be a bit less noisy... Any ideas? I googled for a bit but didn't have any luck.

reader Luboš Motl said...

I am using the Cryosphere Today for all sea-ice related data.


reader AHD said...

Thanks. I don't see any north + south aggregated data, which seems the most natural quantity to follow and care about. But there are links to the data, so I'll put it together myself. Perhaps the antarctic ice volumes are << arctic ice volumes? I believe the antarctic icecap peak area is > that of the arctic icecap, but the arctic icecap is quite a bit thicker I believe. I'll do the analysis and post the answer when I have a chance.

reader Luboš Motl said...

Dear AHD, the global sea ice area is there, too


not to mention that adding two numbers probably isn't that hard. ;-)

reader Howard said...


What is the historical context of CO2 doubling sensitivity? I don't believe there is one as most CO2 changes over the past 2.5Myr are responses (feedbacks) to large changes in climate states.

Another , somewhat related point, is the importance of the application of a forcing and the resulting feedback. If one accepts the M-cycle theory of glacial/interglacial changes, the climate sensitivity is quite high when averaged over the entire globe. Climate scientists get around this by counting ice albedo changes as additional forcing and not a feedback.

My point is that the M-cycle is an example of a concentrated, anisotropic forcing that creates large feedbacks and ultimately flips the climate from warm to ice and vice versa. I call this type of forcing leveraged ("Give me a place to stand, and I shall move the Earth with it").

Man-made well mixed CO2 doubling, on the other hand, is a relatively homogeneous and isotropic forcing. In my earth science stamp collector experience, it seems reasonable that the subsequent feedback mechanisms will be fairly mild and hard to detect.

In addition to CO2, we also have many other man-made changes that can influence climate such as land use changes, irrigated agriculture, carbon black, carbon brown, SOx, NOx, VOCs, dust, ozone creation, ozone destruction, creation of ocean dead zones, jet contrails, etc. Most of these other man-made forcings are heterogeneous anisotropic "leveraged" forcings. Some are positive, some are negative and very little is known about their individual and synergistic feedback mechanisms.

Based on this conceptual model, it seems to me that the actual calculation of CO2 doubling sensitivity requires a detailed understanding of all of these other climate forcings and feedbacks. As far as I can tell, the current IPCC view is that all other man-made forcings even out to zero leaving CO2 as the only driver. This simplistic assumption that makes the sensitivity calculations easy seems "too good to be true" and therefore false.

I appreciates your thoughts and comments, especially if you can point to where I might be full of bullcrap.

reader AHD said...

Whoops, I missed that graph. I still remember my addition, but the graph adds the areas of northern and southern ice of different average thicknesses. The arctic ice is much thicker on average than the antarctic. But perhaps the thinner ice that melts and reforms in the north is of comparable thickness to that in the south. It's certainly thinner than the average across all of the arctic ice. But to gloss over all the details: Isn't the graph that you included a strong counterpoint to the global warming hysteria? The aggregate sea ice shows no odd behavior. It's warmed a bit in the north and cooled a bit in the south. So what?

reader Luboš Motl said...

Of course that the graph is a strong counter-evidence against the climate panic.

All the data in the real world, when taken in a sufficiently broad context - e.g. without picking hemispheres or countries or years - are strong evidence against any climate panic.

reader Luboš Motl said...

Dear Howard, there is no known "historical" - I mean from data from previous geological eras - empirical evidence reliably or semi-reliably supporting the existence of the CO2-related greenhouse effect.

The correlation between CO2 and temperature in the epochs is weak - dominated by other factors that may be said to be "noise" if one assumes the CO2-temperature relationship - and the climate sensitivity per CO2 doubling one computes from those geological data is below 1 Celsius degree, see


There is no evidence that the greenhouse effect played any role in the glaciation cycles. In the glaciation cycles, the concentration of CO2 - as well as all other gases - is driven by the temperature. The temperature is the cause, the trace gases' concentration is a consequence. It can be proven and has been proven by dozens of methods and dozens of papers. If you missed that, sorry, I won't repeat it for you, I've done it about a million of times before.

reader Howard said...


Thanks. I agree that CO2 has not played a detectable role in glacial/interglacial transitions, nor in the DO events as well. Most of the regular scientific papers I have read on this topic admit there is no data supporting a CO2 link.

One slight quibble. The CO2 temperature link is only part of the equation. I believe that changes in ocean circulation and biological activity in the oceans also play a first order role in CO2 variability:


A physiological look at CO2 changes

reader Mark Luhman said...

One of my greatest disappointments in life was when the Democrat controlled congress killed the funding for the supercollider here in the United States, it funny it was Reagan who wanted and the Democrats who killed it and yet here in the US the Republicans are accused of being anti science. The same for nuclear power and Nuclear power research, to the Democrats here in the United States God forbid we study something that may liberate people instead we need to study thing that give them an excuse to regulate people, example AGW.

reader AHD said...

You say: "All the empirical data... are strong evidence against any climate panic". If you're preaching to the choir then no one will argue with that. But all the brainwashed masses have latched onto the arctic sea ice loss as evidence of global warming. This graph is important because it's a simple and compelling demonstration that the arctic sea ice loss is a local phenomenon and that there is no significant global reduction in sea ice.

And I agree with your points about thickness, especially because the thickness of that portion of the polar ice that's seasonal (melts and reforms) is probably comparable in both poles, so no need to complicate the discussion by going from area to volume...

(function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){ (i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o), m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m) })(window,document,'script','//www.google-analytics.com/analytics.js','ga'); ga('create', 'UA-1828728-1', 'auto'); ga('send', 'pageview');