Tuesday 26 July 2011

D0: top forward-backward asymmetry continues to intrigue

While the LHC has been depressingly confirming the predictions of the Standard Model, the good old Tevatron remains the only light in the tunnel. The only 2 lights actually: almost 4 sigma anomaly of the dimuon charge symmetry measured by D0, and over 3 sigma anomaly of the top quark forward-backward asymmetry at high t-tbar invariant mass (yikes) measured by CDF. Future will tell whether these lights should be interpreted as the way out or as the oncoming train. As for the present, the D0 collaboration just provided an important update concerning the forward-backward asymmetry that puts the CDF result in a slightly different... light.

For the latest update D0 used 5.4fb−1 and focused on semi-leptonic top decays. The idea of the measurement is simple: one picks the top decay products, reconstructs the original top and anti-top momenta, and check for an excess of top over antitop quarks moving forward (that is along the proton beam) in the t-tbar rest frame. The Standard Model predicts such an excess should be very small, of order 5%. Instead, after unfolding detector effects from the measured asymmetry, D0 finds the "unfolded" or "production" asymmetry to be 19.6 ± 6.5 %. This kicks in very nicely with the analogous CDF result of 15.8 ± 7.4 % (or 20% when combined with the asymmetry in the dilepton channel). Both results are about 2 sigma away from the Standard Model and both point in the same direction, which is intriguing and almost exciting.

However, not everything agrees perfectly between D0 and CDF. Most importantly, D0 does not see any significant dependence of the asymmetry on the t-tbar invariant mass. On the other hand, CDF sees a dramatic dependence: it was actually the abnormally high asymmetry in the mtt > 450 GeV bin that allowed them to claim a 3-sigma anomaly at the beginning of this year. The situation is thus a bit volatile: the good matching of the inclusive asymmetry between the two experiments is obtained after integrating over discrepant results in the low and high mtt bins.

D0 shares one more important result which I find more exciting. D0 measured the leptonic asymmetry of the leptons originating from top quark decays. The observable is defined as an excess of positive charge leptons moving forward + negative charge leptons moving backward over negative (positive) leptons moving forward (backward). For experimentalists it is more user-friendly than the top asymmetry : it is defined in the laboratory frame and one can avoid tedious and uncertain reconstruction of the momenta of the top quarks. From the theoretical point of view, the lepton asymmetry is tightly related to the top forward-backward asymmetry but not identical. It is related, because the direction of the lepton is clearly correlated with the direction of the mother top or antitop. It is not identical, because that direction is also correlated with the polarization of the mother top. In fact, if the top quark is polarized along some axis, for example in the direction of its motion, the lepton prefers to fly along that direction. See for example this paper for more details. All in all, D0 finds this leptonic asymmetry to be 15.2 ± 4.0%, compared to the Standard Model prediction of 2%. This is more than 3 sigma discrepancy! Not only we get a novel 3 sigma anomaly to cherish, but we also get a hint of anomalous top polarization.

To wrap up , D0 has brought some exciting news and some worrying news too (see the paper for more worries concerning modeling additional QCD radiation that I didn't mention here). The new results will somewhat shake the hierarchy of new physics models that address Tevatron's anomalies, but we have to wait for the next load of theory papers for quantitative details. On the experimental front, next year Tevatron will update all these measurements with twice as much data. About the same time, the LHC will be seriously joining in the game too. Although the LHC cannot measure the top asymmetry directly, due to the symmetric p-p initial state at the LHC, they can access the same physics by constructing more fancy observables. For example, a recent CMS note investigates whether top quarks move closer to the beam axis than anti-top quarks more often than the other way around. Such an effect would be a consequence of a positive forward-backward asymmetry of the t-tbar pair production in quark-antiquark collisions. No effect has been observed in CMS or ATLAS who studied a similar observable. However this doesn't mean much yet: most models addressing Tevatron's top anomaly predict the CMS observable to be the edge of their current sensitivity. It is more worrying that LHC observes no anomalous effects in other top quark observables, like the production cross section or the invariant mass distribution. Yeah, the obstinate lack of new physics at the LHC is utterly worrying. Guys, you better find something fast, otherwise there'll be nothing but darkness.

The D0 paper is available on arXiv. See also Tommaso's comments on the CMS note.


Robert L. Oldershaw said...

Why are you so depressed?

Finally nature is forcing us to question our fundamental assumptions.

The vacuum energy density crisis, the wildly unnatural Planck mass, the lack of an explanation for alpha and h-bar, and all those 26-30 parameters that have to be put into the SM "by hand" have been screaming at us for decades that there is something seriously wrong at the heart of particle physics.

Time for courage and open minds.

In the words of Monty Python: time for something completely different.

And none too soon.


Kea said...

I see that D0 refers to this paper on a Z' explanation with strong quark coupling. Amusingly, the paper is sufficiently recent to hedge its bets about any of the usual Z' suspects.

Anonymous said...

The LHC sees no mass-dependent asymmetry so far (see Monday afternoon t talk)

Manolo said...

Hi Jester,

It looks like D0 does not trust the "coarse" unfolding of CDF, which was employed to arrive at the famous three-sigma anomaly in the asymmetry for mtt>450. CMS does not dare to unfold their mass-dependent results yet, either.

If we forgot that CDF result for a moment (even if it has been kindly feeding theorists these months), things would really look pretty different. D0 (and CMS) results seem to hint to a flat profile in the invariant mass. If that turns out to be true, almost all of the new-physics models that have been proposed so far will be dead (many have been killed already by LHC, anyway), since they predict an asymmetry that increases neatly with mtt.

The one exception I know is our proposal of a very light gluon-like particle, with mass below the t tbar production threshold, exchanged in the s channel. This guy gives a flat (or gently rising) shape of the Tevatron FB asymmetry and of the charge asymmetry at LHC. Moreover, it does not distort the invariant mass distribution of the cross section, neither at Tevatron nor at LHC, so it is almost invisible. I hope some people will investigate this simple idea further.


Jester said...

Right, if there's no mtt dependence of the asymmetry, that'd be a game-changer. But I would not jump to conclusions. A reasonable possibility is that CDF has a 1 sigma upward fluke while D0 has a downward one. A milder mtt dependence actually makes model building easier.

Manolo said...

I agree. That's why I wrote a couple of ifs. (I think the idea of the particle below threshold is nice also for other reasons, but I'm biased.)

JA said...

I'm just waiting to see the new CDF result. I've been told the significance rises. But my spies don't give any more details.

So, I don't think it's a fluke. Perhaps a matter of unfolding.

Anonymous said...

The Nature News article is promoting top quark condensation as a possible explanation. Is this a reasonable interpretation, or did the reporter get spinned by some theorists?

Jester said...

Spinned. I mean, topcolor is a valid model, even if a bit on the baroque side, and all these topgluons and toppions can well produce the asymmetry. But there's nothing in the data that favors this particular model.

Tony Smith said...

As to the 23 July 2011 Nature News article by Ron Cowen saying:
"... Christopher Hill ... proposed how a top quark and its antiparticle could impart mass to the W and Z bosons ...".

Such Higgs as Tquark Condensate structures have been described by Yamawaki and Hashimoto and Tanabashi in papers at
hep-ph/9603293 and hep-ph/0311165
which discuss:

a Bardeen-Hill-Lindner model allows calculation of Higgs mass as around 240 GeV which is in the 250 GeV region of a (statistically small as of now) peak found by both ATLAS and CMS.
Physically, this might correspond to the edge of vacuum stability and the Higgs VEV
a variant (similar to a Kaluza-Klein Nambu-Jona-Lasinio model) allows calculation of Higgs mass as around 176-188 GeV which is near the 205 GeV region of a peak (even smaller statistically) found by both ATLAS and CMS.
Physically, this might correspond to the edge of the Triviality Boundary.