Monday, August 22, 2011

Lepton-Photon Conference Results Show Weaker Higgs Signal


Image from here, shows the Quantum Diaries Survivor plot of what a Standard Model Higgs signal would be expected to look like with the amount of data produced to date for a particular SM Higgs mass values compared to the observations from LHC, as opposed to the usual plot which compared the data to a Standard Model expectation in the absence of a Standard Model Higgs. Thus, at the mid-120 GeVs the signal seen and the signal we would expect to see if it was really there are quite similar, while at higher masses, there is huge gap between the significant of the result we have seen and the significance that we would predict that we would see if it was really there. A Higgs at 125-130 GeV would be expected to produce experimental results much more similar to the ones seen (because experimental power is low in that range producing weak signals in general) than one at say 160 GeV.

The Lepton-Photon Conference this week in Bombay has offered little to encourage scientists hoping that a Standard Model Higgs boson will not be ruled out by experimental evidence.

According to the ATLAS experiment from LHC.

In general, the previous Higgs signals have weakened. Some more details: A 35-minute introductory theoretical talk on the Higgs physics was followed by three 20-minute talks by ATLAS, CMS, and Tevatron. ATLAS - see their fresh new Higgs press release - excluded everything at the 95% level except for 114-146 GeV, 232-256 GeV, 282-296 GeV, and above 466 GeV. No new details about the preferred masses inside these intervals, despite 2/fb of data in some channels. . . .

CMS . . . has mentioned that the p-value for Higgs to gamma-gamma grew deeper, more significant, near 140 GeV but more shallow near 120 GeV. Nothing is seen in the tau-tau channel - which will however grow important in the future. The speaker explains the WW channel ending up with 2l 2nu; then ZZ "golden" channel with 4 leptons. Many combinations are shown. In the CMS p-value, the 120 GeV is actually slightly deeper than for 140 GeV - difference from the gamma-gamma channel itself. Surviving intervals (from the 95% exclusion) are 114-145 GeV, 216-226 GeV, 288-310 GeV, above 340 GeV - different small hills than ATLAS. The deviations grew weaker since the last time. . . . Tevatron has collected 11.5/fb or so and will close on September 30th.

From The Reference Frame.

The more than two sigma deviations from the Standard Model expectation are just barely more than two sigma and aren't always consistent between different experiments. They may not exclude at the 95% confidence level, but they probably do at the 90% confidence level.

One would expect a Standard Model Higgs signal to be more clear at this point (as shown in the image above) and show more of a trendline towards increasing confidence in repeated studies, rather than a persistent low grade, not quite excludable bump in a general area that isn't quite consistent. Instead, any Higgs signal is extremely subtle. And, the studies were looking at are simply looking at the comparison betweeen a null hypothesis and "something" to set their confidence level, not a relative hypothesis test between some highly specific predictation and the results seen. At this point the Higgs boson prediction has degenerated from a specific spin zero particle at a fairly specific mass (that has already been ruled out and is close to the Z boson mass) to a prediction that there is something like it out there somewhere.

The "width" of a signal in one of these charts is a measure of its "cross-section" which is quite narrow for all particles in the Standard Model discovered to date. Nothing observed so far has such a shallow, wide bump as opposed to a steep narrow bump. If what Atlas and CMS are seeing as a deviation from the Standard Model is an undiscovered single particle of any kind, it is a very different animal than anything else seen to date or what theorists had expected a Higgs boson to look like (perhaps in ignorance of how a spin-0 fundmamental particle differs from either a higher spin fermion or boson, or a spin-0 meson which is composite).]

Precision electroweak data fitted to the Standard Model and theoretical considerations in models like the mimially supersymmetric standard model (the most famous of the SUSY models) strongly disfavor any of the mass ranges for the Higgs except the 114-145 GeV range that still remains standing, barely. The Higgs boson mass range favored by precision electroweak data of ca. 90-110 GeV has already been ruled out.

Marco Frasca is pretty much ready to write off the Standard Model Higgs boson in favor of "a strongly coupled Higgs boson" within the context of a SUSY model, as he explains in technical detail here. In the sense used "strongly coupled" mere means having a very large, self-interacting coupling constant, rather than specifically having a strong force of QCD coupling through gluons, although this Higgs mass generation mechanism is an analogy to the way that mass is generated in QCD through strong force interactions.

Woit sums it up thusly:

•No Higgs above 145 GeV
•In the region 135-145 GeV, both experiments are seeing somewhat more events than expected from background, but less than expected if there really was a Higgs there.
•Not enough data to say anything about 115-135 GeV, the Higgs could still be hiding there. If so, a malicious deity has carefully chosen the Higgs mass to make it as hard as possible for physicists to study it.
.

Particle v. Antiparticle Mass

Meanwhile, from the same link:

In recent 3 days, the ASACUSA experiment at CERN announced its accurate laser measurements of the antiproton mass. It agrees with the proton mass at the accuracy of 1 part per billion. The CPT symmetry which almost certainly holds exactly implies that matter and antimatter - whenever they can be distinguished by charges to be sure that you made the "anti-" operation correctly - have exactly the same mass. . . .

Just compare the measurement of the antiproton mass with some of the spectacular claims about the top-antitop mass difference. It's almost the same thing but CDF at the Tevatron claimed that the antitop mass differs from the top mass by 3 GeV or so - by two percent.

Try to invent a reason why the top and antitop masses would differ by 2% - but the analogous proton-antiproton mass difference (which may be affected by the top-antitop differences as well) would be smaller by more than 7 orders of magnitude. I don't say that it's strictly impossible to invent such a mechanism but it is probably going to be very awkward.

Claims that someone observes CPT violation are extraordinary claims and they require extraordinary evidence. So if one measures the top-antitop mass difference to be nonzero at a 2-sigma or 3-sigma level (by 3 GeV), he should say that "our measurements were rather inaccurate; our observation of the mass difference had the error of 3 GeV".

On one hand, I don't entirely agree with Lubos Motl (I'd put the accent mark in there if I knew my keystrokes better), in the quote above, that it is so hard theoretically to know that proton and anti-proton mass are within one part per billion of each other (10^-9), while the top and antitop pair differ by two parts per hundred (2x10^-2). The top quark weighs 174,000 MeV/c^2 while is first generation up and down quarks that go into a proton are almost five orders of magnitude lighter each. And, in a proton more than 90% of the mass comes from the strong nuclear force that binds the two up and one down quarks in it together, not from the quarks themselves. Thus, there any mass difference signal is damped by about six orders of magnitude in a proton relative to a top quark and is further damped by the need to disentangle up and down quark mass differences in a proton, when those masses aren't known with as much precision as we'd like since quark confinement and the relative stability of protons and neutrons made of them makes it harder to measure them separately from each other. So it isn't entirely unreasonable to see that level of experimental error in the two systems as close to comparable in some respects.

The fact that top quarks almost exclusively decay first into bottom quarks, without forming hadrons (three quark baryons or two quark mesons) as other quarks do, also could make the signal cleaner and hence make it easier to get a mass difference to appear in these decays that would be hard to seen in particles that have error inducing background from QCD hadron formation possibilites.

On the other hand, the theoretical and experimental motivations for particles and their antiparticles to have the same masses is extremely strong. And there is a huge range of masses between the top quark scale and the components quarks of protons where particle-antiparticle mass equivalency has been confirmed to considerable experimental precision. Theoretically, the need for this symmetry goes all the way back to Dirac's equation and has remains central in all subsequent elaborations of quantum mechanics. So, Motl appropriately invokes the extraordinary claims require extraordinary evidence maxim.

The only person I'm aware of who is real touting a model that entertains this kind of asymmetry is the scientifically trained physics blogger Kea from New Zealand. But, while those models are interesting, I'm not yet convinced that they are sufficiently well motivated to back the existence of a CPT violation supported by only a modest two or three sigma discrepency in pretty much one experiment.

Personal Speculation Regarding Composite Quarks

The attraction of Higgsless models and heavy Higgs models, relative to more traditional Standard Model and SUSY variations, on it continues to surge.

For example, one quite natural way to get mass with fundamentally massless fundamental particles, would be to assume that quarks are composite.

The analogy to the atomic nucleus to proton and neutron to quark relationship could be quite compelling. The nuclear binding force between protons and neutrons is mediated by a composite meson (which is also a boson) called the pion and this effective nuclear binding force is essentially overflow from the strong force interactions between the quarks in individual protons and neutrons mediated by gluons. Gluons are formally fundamental in the Standard Model, but the fact that each gluon must have two color charges, one of which must be an anti-color charge, and exchange exchange color charges between quarks, the model is just a hair short of a composite one already.

The very limited amount of direct observation of the inner workers of protons and neutrons and mesons due to quark confinement, and the fact that QCD predictions from first principles tend to match real world observations only at about two significant digit accuracy, despite its best efforts, leaves a significant gap between the strongly theoretically and experimentally motivated QCD rules we take to be true in the Standard Model and what we have actually confirmed to any precision experimentally. For example, the experimental limitations that compel the gluon to be actually massless, rather than merely having a very low mass (perhaps on the order of a neutrino), are quite weak since the strong force operates at such short ranges.

It would not be all that earth shaking to discover that quarks and gluons are themselves composite and that the strong force is merely a spill over of a preon binding force that holds preons together in a quark or gluon. The natural energy level of such a confining force would be immense rendering this binding force a natural source for all or most of the mass in a composite quark, just as binding force is responsible for 90% of the mass of a proton or neutron and a significant and measurable amount of the nuclear mass of an atom. Even if these preons had some fundamental mass, it would be easy to imagine that this mass was on the same order of magnitude as the leptons, with the rest arising from a binding force.

Preon models aren't that hard to devise, and some fairly good ones have been developed, although it is harder for those models to explain in any way the three generations of fermions that are observed, experimental evidence that would prove those models that is within the realm of possibility, or the way that the weak force decay probabilities seem to put different fundamental particles in the Standard Model on equal footing (although even then, the fact that the weak force treats each color of quark as a distinct possibility relative to a single charged lepton possibility at each generation can be seen as an argument that quarks but not leptons are composite).

It also isn't too hard to imagine that the three color charges might correspond to the three spatial dimensions, while the electrical charge might correspond to the time dimension in some profound way. In this view, the relative strength of color force relative to electrical charge would be related to the fact that the conversion factor of the speed of light relates very large spatial distances to very small time distances. The fact that both color force and electromagnetic force appear to be non-CP violating would also suggest a link between the two. Now, this program of unification would leave the weak force, which has been unified with the electromagnetic force, the odd man out after decades of wedded bliss with electromagnetism, but surely some sense of its role could emerge from this line of thought.

Experimental limitations on composite quarks and gluons also ease up quite a bit if you assume that the effects expects at the Planck scale actually manifest several orders of magnitude smaller than the Planck length.

2 comments:

Anonymous said...

Lubos is right, CPT invariance is fundamental and Fermi lab was negligent in its publication. As for preon models, they imply similar problems with the symmetry, and are unlikely for the same reason.

Andrew Oh-Willeke said...

It isn't obvious to me what symmetry problems arise from preons.

I'd also disagree that Fermi lab is negligent in publishing any results of its experiments. Arguably a different interpretation showing greater skepticism would be in order, but you should publish whatever you find, especially when you are the only game in town.