Friday, 1 April 2016

April Fools' 16: Was LIGO a hack?


This post is an April Fools' joke. LIGO's gravitational waves are for real. At least I hope so ;) 

We have had recently a few scientific embarrassments, where a big discovery announced with great fanfares was subsequently overturned by new evidence.  We still remember OPERA's faster than light neutrinos which turned out to be a loose cable, or BICEP's gravitational waves from inflation, which turned out to be galactic dust emission... It seems that another such embarrassment is coming our way: the recent LIGO's discovery of gravitational waves emitted in a black hole merger may share a similar fate. There are reasons to believe that the experiment was hacked, and the signal was injected by a prankster.

From the beginning, one reason to be skeptical about LIGO's discovery was that the signal  seemed too beautiful to be true. Indeed, the experimental curve looked as if taken out of a textbook on general relativity, with a clearly visible chirp signal from the inspiral phase, followed by a ringdown signal when the merged black hole relaxes to the Kerr state. The reason may be that it *is* taken out of a  textbook. This is at least what is strongly suggested by recent developments.

On EvilZone, a well-known hacker's forum, a hacker using a nickname Madhatter was boasting that it was possible to tamper with scientific instruments, including the LHC, the Fermi satellite, and the LIGO interferometer.  When challenged, he or she uploaded a piece of code that allows one to access LIGO computers. Apparently, the hacker took advantage the same backdoor that allows the selected members of the LIGO team to inject a fake signal in order to test the analysis chain.  This was brought to attention of the collaboration members, who  decided to test the code. To everyone's bewilderment, the effect was to reproduce exactly the same signal in the LIGO apparatus as the one observed in September last year!

Even though the traces of a hack cannot be discovered, there is little doubt now that there was a foul play involved. It is not clear what was the motif of the hacker: was it just a prank, or maybe an elaborate plan to discredit the scientists. What is even more worrying is that the same thing could happen in other experiments. The rumor is that the ATLAS and CMS collaborations are already checking whether the 750 GeV diphoton resonance signal could also be injected by a hacker.

Thursday, 17 March 2016

Diphoton update

Today at the Moriond conference ATLAS and CMS updated their diphoton resonance searches. There's been a rumor of an ATLAS analysis with looser cuts on the photons where the significance of the 750 GeV excess grows to a whopping 4.7 sigma. The rumor had it that the this analysis would be made public today, so the expectations were high. However, the loose-cuts analysis was not approved in time by the collaboration, and the fireworks display was cancelled.  In any case,  there was some good news today, and some useful info for model builders was provided.











Let's start with ATLAS. For the 13 TeV results, they now have two analyses: one called spin-0 and one called spin-2. Naively, the cuts in the latter are not optimized not for a spin-2 resonance but rather for a high-mass resonance  (where there's currently no significant excess), so the spin-2 label should not be treated too seriously in this case. Both analyses show a similar excess at 750 GeV: 3.9 and 3.6 sigma respectively for a wide resonance. Moreover, ATLAS provides additional information about the diphoton events, such as the angular distribution of the photons, the number of accompanying jets, the amount of missing energy, etc. This may be very useful for theorists entertaining less trivial models, for example when the 750 GeV resonance is produced  from a decay of a heavier parent particle. Finally, ATLAS shows a re-analysis of the diphoton events collected at 8 TeV center-of-energy of the LHC. The former run-1 analysis was a bit sloppy in the interesting mass range; for example, no limits at all were given for a 750 GeV scalar hypothesis.  Now the run-1 data have been cleaned up and analyzed using the same methods as in run-2. Excitingly, there's a 2 sigma excess in the spin-0 analysis in run-1, roughly compatible with what one would expect given the observed run-2 excess!   No significant excess is seen for the spin-2 analysis, and the tension between the run-1 and run-2 data is quite severe in this case. Unfortunately, ATLAS does not quote the combined significance and the best fit cross section for the 750 GeV resonance.

For CMS, the big news is that the amount of 13 TeV data at their disposal has increased by 20%. Using MacGyver skills, they managed to make sense of the chunk of data collected when the CMS magnet was off due to a technical problem. Apparently it was worth it, as new diphoton events have been found in the 750 GeV ballpark. Thanks to that, and a better calibration,  the significance of the diphoton excess in run-2  actually increases up to 2.9 sigma!  Furthermore, much like ATLAS, CMS updated their run-1 diphoton analyses and combined them with the run-2 ones.  Again, the combination increases the significance of the 750 GeV excess. The combined significance quoted by CMS is 3.4 sigma,  similar for spin-0 and spin-2 analyses. Unlike in ATLAS, the best fit is for a narrow resonance, which is the more preferred option from the theoretical point of view.

In summary, the diphoton excess survived the first test.  After adding more data and improving the analysis techniques the significance slightly increases rather than decreases, as expected for a real particle.  The signal is now a bit more solid: both experiments have a similar amount of diphoton data and they both claim a similar significance of the  750 GeV bump.  It may be a good moment to rename the ATLAS diphoton excess as the LHC diphoton excess :)  So far, the story of 2012 is repeating itself: the initial hints of a new resonance solidify into a consistent picture. Are we going to have another huge discovery this summer?

Friday, 11 March 2016

750 GeV: the bigger picture

This Thursday the ATLAS and CMS experiments will present updated analyses of the 750 GeV diphoton excess. CMS will extend their data set by the diphoton events collected in the periods when the detector was running without the magnetic field (which is not essential for this particular study), so the amount of available data will slightly increase. We will then enter the Phase-II of the excitement protocol,  hopefully followed this summer by another 4-th-of-July-style discovery. To close the Phase-I, here's a long-promised post about the bigger picture. There's at least 750 distinct models that can accommodate the diphoton signal observed by ATLAS and CMS. However, a larger framework for physics beyond the Standard Model it which these phenomenological models can be embedded is a more tricky question. Here is a bunch of speculations.

Whenever a new fluctuation is spotted at the LHC one cannot avoid mentioning supersymmetry. However,  the 750 GeV resonance cannot be naturally interpreted in this framework, not the least because it cannot be identified as a superpartner of any known particles. The problem is that explaining the observed signal strength requires introducing new particles with large couplings, and the complete theory typically enters into a strong coupling regime at the energy scale of a few TeV. This is not the usual SUSY paradigm, with weakly coupled physics at the TeV scale followed by a desert up to the grand unification scale. Thus, even if the final answer may still turn out to be supersymmetric, it will not be the kind of SUSY we've been expecting all along. Weakly coupled supersymmetric explanations are still possible in somewhat more complicated scenarios with new very light sub-GeV particles and cascade decays, see e.g. this NMSSM model.

Each time you see a diphoton peak you want to cry Higgs, since this is how the 125 GeV Higgs boson was first spotted. Many theories predict an extended Higgs sector with multiple heavy scalar particles, but again such a framework is not the most natural one for interpreting the 750 GeV resonance. There are two main reasons. One is that different Higgs scalars typically mix, but the mixing angle in this case is severely constrained by Higgs precision studies and non-observation of 750 GeV diboson resonances in other channel. The other is that, for a 750 GeV Higgs scalar, the branching fraction into the diphoton final state is typically tiny (e.g., ~10^-7 for a Standard-Model-Higgs-like scalar) and a complicated model gymnastics is needed to enhance it. The possibility that the 750 GeV resonance is a heavy Higgs boson is by no means excluded, but I would be surprised if this were the case.  

It is more tempting to interpret the diphoton resonance as a bound state of new strong interactions with a confinement scale in the TeV range. We know that the Quantum Chromodynamics (QCD) theory, which describes the strong interactions of the Standard Model quarks, gives rise to many scalar mesons and higher-spin resonances at low energies. Such a behavior is characteristic for a large class of similar theories.  Furthermore,  if the new strong sector contains mediator particles  that carry color and electromagnetic charges, the production in gluon fusion and decay into photons is possible for the composite states, see e.g. here.  The problem is that, much as for QCD, one would expect not one but an entire battalion of resonances. One needs to understand how the remaining resonances predicted by typical strongly interacting models could have avoided detection so far.

One way this could happen is if the 750 GeV resonance is a scalar that, for symmetry reasons, is much lighter than most of the particles in the strong sector. Here again our QCD may offer us a clue, as it contains pseudo-scalar particles, the so-called pions,  which are a factor of 10 lighter than the typical mass scale of other resonances. In QCD, pions are Goldstone bosons of the chiral symmetry spontaneously broken by the vacuum quark condensate. In other words, the smallness of the pion mass is  protected by a symmetry, and general theorems worked out in the 60s  ensure the quantum stability of such an arrangement. The similar mechanism can be easily implemented in other strongly interacting theories,  and it is possible to realize the 750 GeV resonance as a new kind of pion, see e.g. here.   Even the mechanism for decaying into photons -- via chiral anomalies -- can be borrowed directly from QCD. However, the symmetry protecting the 750 GeV scalar could also be completely different that the ones we have seen so far. One example is the dilaton, that is   a Goldstone boson of a spontaneously broken conformal symmetry, see e.g. here. This is a theoretically interesting possibility, since approximate conformal symmetry often arises as a feature of strongly interacting theories. All in all, the 750 GeV particle may well be  a pion or dilaton harbinger of new strong interactions at a TeV scale. One can then further speculate that the Higgs boson also originates from that sector, but that is a separate story that may or may not be true.

Another larger framework worth mentioning here is that of extra dimensions. In the modern view, theories with the new 4th dimension of space are merely an effective description of strongly interacting sectors discussed above. For example, the famous Randall-Sundrum model, with the Standard Model living in a section of a 5D AdS5 space, is a weakly coupled dual description of strongly coupled theories with a conformal symmetry and a large N gauge symmetry. These models thus offer a calculable way to embed the 750 GeV resonance in a strongly interacting theory. For example, the dilaton can be effectively described in the Randall-Sundrum model as the radion - a scalar particle corresponding to fluctuations of the size of the 5th dimension, see e.g. here. Moreover, the Randall-Sundrum framework  provides a simple way to realize the 750 GeV particle as a spin-2 resonance. Indeed, the model always contains massive Kaluza-Klein  excitations of the graviton, whose couplings to matter can be much stronger than that of the massless graviton. This possibility have been relatively less explored so far, see e.g.  here,  but that may change next week...

Clearly, it is impossible to say anything conclusive at this point. More data in multiple decay channels is absolutely necessary  for a more concrete picture to emerge. For me personally, a confirmation of the 750 GeV excess would be a strong hint for new strong interactions at a few TeV scale. And if this is indeed the case,  one may seriously think that our  40-years-long brooding about the hierarchy problem has not been completely misguided...

Thursday, 11 February 2016

LIGO: what's in it for us?

I mean us theoretical particle physicists. With this constraint, the short answer is not much.  Of course, every human being must experience shock and awe when picturing the phenomenon observed by LIGO. Two black holes spiraling into each other and merging in a cataclysmic event which releases energy equivalent to 3 solar masses within a fraction of a second.... Besides, one can only  admire the ingenuity that allows us to detect here on Earth a disturbance of the gravitational field created in a galaxy 1.3 billion light years away. In more practical terms, the LIGO announcement starts the era of gravitational wave astronomy and thus opens a new window on the universe. In particular, LIGO's discovery is a first ever observation of a black hole binary, and we should soon learn more about the ubiquity of astrophysical systems containing one or more black holes. Furthermore, it is possible that we will discover completely new objects whose existence we don't even suspect. Still, all of the above is what I fondly call dirty astrophysics on this blog,  and it does not touch upon any fundamental issues. What are the prospects for learning something new about those?

In the long run, I think we can be cautiously optimistic. While we haven't learned anything unexpected from today's LIGO announcement, progress in gravitational wave astronomy should eventually teach us something about fundamental physics. First of all, advances in astronomy, inevitably brought by this new experimental technique,  will allow us to better measure the basic parameters of the universe. This in turn will provide us information about aspects of fundamental physics that can affect the entire universe, such as e.g. the dark energy. Moreover, by observing phenomena occurring in strong gravitational fields and of which signals propagate over large distances, we can place constraints on modifications of Einstein gravity such as the graviton mass (on the downside,  often there is no consistent alternative theory that can be constrained).

Closer to our hearts, one potential source of gravitational waves is a strongly first-order phase transition. Such an event may have occurred as the early universe was cooling down.  Below a certain critical temperature a symmetric phase of the high-energy theory may no longer be energetically preferred, and the universe enters a new phase where the symmetry is broken. If the transition is violent (strongly first-order in the physics jargon), bubbles of the new phase emerge, expand, and collide, until they fill the entire visible universe. Such a dramatic event produces gravitational waves with the amplitude that may be observable by future experiments.   Two examples of phase transitions we suspect to have occurred are the QCD phase transition  around T=100 MeV, and the electroweak phase transition around T=100 GeV. The Standard Model predicts that neither is first order, however new physics beyond the Standard Model may change that conclusion. Many examples of required  new physics have been proposed to modify the electroweak phase transition, for example models with additional Higgs scalars, or with warped extra dimensions.  Moreover, the phase transition could be related to symmetry breaking in a hidden sector that is very weakly or not at all coupled (except via gravity) to ordinary matter.  Therefore, by observing or putting limits on phase transitions in the early universe we will obtain complementary information about the fundamental theory at high energies.    

Gravitational waves from phase transitions are typically predicted to peak at frequencies much smaller than the ones probed by LIGO (35 to 250 Hz). The next generation of gravitational telescopes will be more equipped to detect such a signal thanks to a much larger arm-length (see figure borrowed from here). This concerns especially the eLISA space interferometer which will probe millihertz frequencies. Even smaller frequencies can be probed by pulsar timing arrays which search for signals of gravitational waves using stable pulsars for an antenna.  The worry is that the  interesting signal may be obscured by astrophysical backgrounds, such as (oh horror) gravitational wave emission from white dwarf binaries. Another  interesting beacon for future experiments is to detect gravitational waves from inflation (almost discovered 2 years ago via another method by the BICEP collaboration).  However, given the constraints from the CMB observations,  the inflation signal may well be too weak even for the future  giant space interferometers like DECIGO or BBO.

To summarize, the importance of the LIGO discovery for the field  of particle physics is mostly the boost it gives to  further experimental efforts in this direction.  Hopefully, the eLISA project will now take off, and other ideas will emerge. Once gravitational wave experiments become sensitive to sub-Hertz frequencies, they will start probing the parameter space of interesting theories beyond the Standard Model.  

Thanks YouTube! It's the first time I see a webcast of a highly anticipated event running smoothly in spite of 100000 viewers. This can be contrasted with Physical Review Letters who struggled to make one damn pdf file accessible ;) 

Sunday, 31 January 2016

750 ways to leave your lover

A new paper last week straightens out the story of the diphoton background in ATLAS. Some confusion was created because theorists misinterpreted the procedures described in the ATLAS conference note, which could lead to a different estimate of the significance of the 750 GeV excess. However, once the correct phenomenological and statistical approach is adopted, the significance quoted by ATLAS can be reproduced, up to small differences due to incomplete information available in public documents. Anyway, now that this is all behind, we can safely continue being excited at least until summer.  Today I want to discuss different interpretations of the diphoton bump observed by ATLAS. I will take a purely phenomenological point of view, leaving for the next time  the question of a bigger picture that the resonance may fit into.

Phenomenologically, the most straightforward interpretation is the so-called everyone's model: a 750 GeV singlet scalar particle produced in gluon fusion and decaying to photons via loops of new vector-like quarks. This simple construction perfectly explains all publicly available data, and can be easily embedded in more sophisticated models. Nevertheless, many more possibilities were pointed out in the 750 papers so far, and here I review a few that I find most interesting.

Spin Zero or More?  
For a particle decaying to two photons, there is not that many possibilities: the resonance has to be a boson and, according to young Landau's theorem, it cannot have spin 1. This leaves at the table spin 0, 2, or higher. Spin-2 is an interesting hypothesis, as this kind of excitations is predicted in popular models like the Randall-Sundrum one. Higher-than-two spins are disfavored theoretically. When more data is collected, the spin of the 750 GeV resonance can be tested by looking at the angular distribution of the photons. The rumor is that the data so far somewhat favor spin-2 over spin-0, although the statistics is certainly insufficient for any serious conclusions.  Concerning the parity, it is practically impossible to determine it by studying the diphoton final state, and both the scalar and the pseudoscalar option are equally viable at present. Discrimination may be possible in the future, but  only if multi-body decay modes of the resonance are discovered. If the true final state is more complicated than two photons (see below), then the 750 GeV resonance may have  any spin, including spin-1 and spin-1/2.

Narrow or Wide? 
The total width is an inverse of particle's lifetime (in our funny units). From the experimental point of view, the width larger than detector's  energy resolution  will show up as a smearing of the resonance due to the uncertainty principle. Currently, the ATLAS run-2 data prefer the width 10 times larger than the experimental resolution  (which is about 5 GeV in this energy ballpark), although the preference is not very strong in the statistical sense. On the other hand, from the theoretical point of view, it is much easier to construct models where the 750 GeV resonance is a narrow particle. Therefore, confirmation of the large width would have profound consequences, as it would significantly narrow down the scope of viable models.  The most exciting interpretation would then be that the resonance is a portal to a dark sector containing new light particles very weakly coupled to ordinary matter.    

How many resonances?  
One resonance is enough, but a family of resonances tightly packed around 750 GeV may also explain the data. As a bonus, this could explain the seemingly large width without opening new dangerous decay channels. It is quite natural for particles to come in multiplets with similar masses: our pion is an example where the small mass splitting π± and π0 arises due to electromagnetic quantum corrections. For Higgs-like multiplets the small splitting may naturally arise after electroweak symmetry breaking, and  the familiar 2-Higgs doublet model offers a simple realization. If the mass splitting of the multiplet is larger than the experimental resolution, this possibility can tested by precisely measuring the profile of the resonance and searching for a departure from the Breit-Wigner shape. On the other side of the spectrum is the idea is that there is no resonance at all at 750 GeV, but rather at another mass, and the bump at 750 GeV appears due to some kinematical accidents.
   
Who made it? 
The most plausible production process is definitely the gluon-gluon fusion. Production in collisions of light quark and antiquarks is also theoretically sound, however it leads to a more acute tension between run-2 and run-1 data. Indeed, even for the gluon fusion, the production cross section of a 750 GeV resonance in 13 TeV proton collisions is only 5 times larger than at 8 TeV. Given the larger amount of data collected in run-1, we would expect a similar excess there, contrary to observations. For a resonance produced from u-ubar or d-dbar the analogous ratio is only 2.5 (see the table), leading to much more  tension. The ratio climbs back to 5 if the initial state contains the heavier quarks: strange, charm, or bottom (which can also be found sometimes inside a proton), however I haven't seen yet a neat model that makes use of that. Another possibility is to produce the resonance via photon-photon collisions. This way one could cook up a truly minimal and very predictive model where the resonance couples only to photons of all the Standard Model particles. However, in this case, the ratio between 13 and 8 TeV cross section is very unfavorable, merely a factor of 2, and the run-1 vs run-2 tension comes back with more force. More options open up when associated production (e.g. with t-tbar, or in vector boson fusion) is considered. The problem with these ideas is that, according to what was revealed during the talk last December, there isn't any additional energetic particles in the diphoton events. Similar problems are facing models where the 750 GeV resonance appears as a decay product of a heavier resonance, although in this case some clever engineering or fine-tuning may help to hide the additional particles from experimentalist's eyes.

Two-body or more?
While a simple two-body decay of the resonance into two photons is a perfectly plausible explanation of all existing data, a number of interesting alternatives have been suggested. For example, the decay could be 3-body, with another soft visible or invisible  particle accompanying two photons. If the masses of all particles involved are chosen appropriately, the invariant mass spectrum of the diphoton remains sharply peaked. At the same time, a broadening of the diphoton energy due to the 3-body kinematics may explain why the resonance appears wide in ATLAS. Another possibility is a cascade decay into 4 photons. If the  intermediate particles are very light, then the pairs of photons from their decay are very collimated and may look like a single photon in the detector.
   
 ♬ The problem is all inside your head   and the possibilities are endless. The situation is completely different than during the process of discovering the  Higgs boson, where one strongly favored hypothesis was tested against more exotic ideas. Of course, the first and foremost question is whether the excess is really new physics, or just a nasty statistical fluctuation. But if that is confirmed, the next crucial task for experimentalists will be to establish the nature of the resonance and get model builders on the right track.  The answer is easy if you take it logically ♬ 

All ideas discussed above appeared in recent articles by various authors addressing the 750 GeV excess. If I were to include all references the post would be just one giant hyperlink, so you need to browse the literature yourself to find the original references.

Friday, 22 January 2016

Higgs force awakens

The Higgs boson couples to particles that constitute matter around us, such as electrons, protons, and neutrons. Its virtual quanta are constantly being exchanged between these particles.  In other words, it gives rise to a force -  the Higgs force. I'm surprised why this PR-cool aspect is not explored in our outreach efforts. Higgs bosons mediate the Higgs force in the same fashion as gravitons, gluons, photons, W and Z bosons mediate  the gravity, strong, electromagnetic, and  weak forces. Just like gravity, the Higgs force is always attractive and its strength is proportional, in the first approximation, to particle's mass. It is a force in a common sense; for example, if we bombarded long enough a detector with a beam of particles interacting only via the Higgs force, they would eventually knock off atoms in the detector.

There is of course a reason why the Higgs force is less discussed: it has never been detected directly. Indeed, in the absence of midi-chlorians it is extremely weak. First, it shares the feature of the weak interactions of being short-ranged: since the mediator is massive, the interaction strength is exponentially suppressed at distances larger than an attometer (10^-18 m), about 0.1% of the diameter of a proton. Moreover, for ordinary matter, the weak force is more important because of the tiny Higgs couplings to light quarks and electrons. For example, for the proton the Higgs force is thousand times weaker than the weak force, and for the electron it is hundred thousand times weaker. Finally, there are no known particles interacting only via the Higgs force and gravity (though dark matter in some hypothetical models has this property), so in practice the Higgs force is always a tiny correction to more powerful forces that shape the structure of atoms and nuclei. This is again in contrast to the weak force, which is particularly relevant for neutrinos who are immune to strong and electromagnetic forces.

Nevertheless, this new paper argues that the situation is not hopeless, and that the current experimental sensitivity is good enough to start probing the Higgs force. The authors propose to do it by means of atom spectroscopy. Frequency measurements of atomic transitions have reached the stunning accuracy of order 10^-18. The Higgs force creates a Yukawa type potential between the nucleus and orbiting electrons, which leads to a shift of the atomic levels. The effect is tiny, in particular it  is always smaller than the analogous shift due to the weak force. This is a serious problem, because calculations of the leading effects may not be accurate enough to extract the subleading Higgs contribution.  Fortunately, there may be tricks to reduce the uncertainties. One is to measure how the isotope shift of transition frequencies for several isotope pairs. The theory says that the leading atomic interactions should give rise to a universal linear relation (the so-called King's relation) between  isotope shifts for different transitions. The Higgs and weak interactions should lead to a violation of King's relation. Given many uncertainties plaguing calculations of atomic levels, it may still be difficult to ever claim a detection of the Higgs force. More realistically, one can try to set limits on the Higgs couplings to light fermions which will be better than the current collider limits.  

Atomic spectroscopy is way above my head, so I cannot judge if the proposal is realistic. There are a few practical issues to resolve before the Higgs force is mastered into a lightsaber. However, it is possible that a new front to study the Higgs boson will be opened in the near future. These studies will provide information about the Higgs couplings to light Standard Model fermions, which is complementary to the information obtained from collider searches.

Sunday, 17 January 2016

Gunpowder Plot: foiled

Just a week ago I hailed the new king, and already there was an assassination attempt. A new paper claims that the statistical significance of the 750 GeV diphoton excess is merely 2 sigma local. The  story is being widely discussed in the corridors and comment sections because we all like to watch things die...  The assassins used this plot:

The Standard Model prediction for the diphoton background at the LHC is difficult to calculate from first principles. Therefore,  the ATLAS collaboration assumes a theoretically motivated functional form for this background as a function of the diphoton invariant mass. The ansatz contains a number of free parameters, which are then fitted using the data in the entire analyzed range of invariant masses. This procedure leads to the prediction represented by the dashed line in the plot (but see later). The new paper assumes a slightly more complicated functional form with more free parameters, such that the slope of the background is allowed to change.  The authors argue that their more general  ansatz provides a better fit to the entire diphoton spectrum, and moreover predicts a larger background for the large invariant masses.  As a result, the significance of the 750 GeV excess decreases to an insignificant value of 2 sigma.
     
There are several problems with this claim.  First, I'm confused why the blue line is described as the ATLAS fit, since it is clearly different than the background curve in the money-plot provided by ATLAS (Fig. 1 in ATLAS-CONF-2015-081). The true ATLAS background is above the blue line, and much closer to the black line in the peak region (edit: it seems now that the background curve plotted by ATLAS corresponds to a1=0  and one more free parameter for an overall normalization, while the paper assumes fixed normalization). Second, I cannot reproduce the significance quoted in the paper. Taking the two ATLAS bins around 750 GeV, I find 3.2 sigma excess using the true ATLAS background, and 2.6 sigma using the black line (edit: this is because my  estimate is too simplistic, and the paper also takes into account the uncertainty on the background curve). Third, the postulated change of slope is difficult to justify theoretically. It would mean there is a new background component kicking in at ~500 GeV, but this does not seem to be the case in this analysis.

Finally, the problem with the black line is that it grossly overshoots the high mass tail,  which is visible even to a naked eye.  To be more quantitative, in the range 790-1590 GeV there are 17 diphoton events observed by ATLAS,  the true ATLAS backgrounds predicts 19 events, and the black line predicts 33 events. Therefore, the background shape proposed in the paper is inconsistent with the tail at the 3 sigma level! While the alternative background choice decreases the  significance at the 750 GeV peak, it simply moves (and amplifies) the tension to another place.

So, I think the plot is foiled and the  claim does not stand scrutiny.  The 750 GeV peak may well be just a statistical fluctuation that will go away when more data is collected, but it's unlikely to be a stupid error on the part of ATLAS. The king will live at least until summer.