Saturday, 5 March 2011

CDF: curiouser and curiouser

The Tevatron may be a dead man walking but it continues to kick ass. The CDF collaboration just posted a new measurement of the forward-backward asymmetry of the top pair production. Recall that earlier this year CDF made a surprising claim about that asymmetry. Restricting to the top pairs with the invariant mass larger than 450 GeV (about 30% of all t-tbar events) the asymmetry is stunning 48±11 percent, which is 3.4 sigma away from the Standard Model prediction of 9 percent. This is completely crazy: 3 times more often top quarks choose to shoot forward rather than backward (with respect to the direction of the proton beam), even though in the first approximation they should not prefer any direction. Even fancy new physics model have a hard time to predict such a huge asymmetry.

The previous CDF measurement was dealing with semileptonic top decays when one top or anti-top quark decays leptonically to an electron/muon + neutrino + b-quark, while the other decays to 3 quarks. The new measurement that I'm reviewing here is focused on dileptonic decays when both the top and the anti-top decay leptonically. This type of event is more rare; only about 5% of the top pairs decay in this manner. Nevertheless, the top stash at the Tevatron is now large enough. In 5.1 inverse femtobarn that went into the new analysis one expects over 200 dileptonic top events. This is enough to study differential distributions and asymmetries. The CDF note gives the result for the inclusive forward-backward asymmetry, that is for the entire dileptonic sample regardless of the invariant mass of the reconstructed top pair. The measured inclusive asymmetry is 14± 5% which, after unfolding the background and instrumental effects, corresponds to the parton level asymmetry of 42 ± 16 percent. The Standard Model predicts meager 5% so the discrepancy is 2.3 sigma. Much as in the semileptonic sample, the asymmetry is larger at higher t-tbar invariant masses (see the pictures), however poor statistics precludes any firm conclusions. One should also compare that result to the inclusive asymmetry of the semileptonic sample. The latter is much smaller, 16±7%, nevertheless the two results are consistent within 2 sigma.

Formally, the new CDF result is merely a 2 sigma deviation from the Standard Model. However, when combined with the previous 3 sigma anomaly, it has a much stronger psychological impact. One could worry that the CDF measurement of the asymmetry suffers from some unaccounted for systematic effects. In fact, the semileptonic sample has a quirky trait that the entire asymmetry comes from the events featuring a muon, while the events containing an electron do not show a significant asymmetry. Until very recently the following explanations of the anomaly seemed equally plausible:
  • a cat got stranded in the CDF muon chambers,
  • the QCD contribution to the asymmetry has been underestimated,
  • the asymmetry is a manifestation of new physics.
The new result makes the cat hypothesis less likely. In the dileptonic sample the asymmetry is actually the largest for the dielectron events. The systematic effects are quite different for the two measurements, yet both consistently show a large positive asymmetry, which is reassuring.

Now, it remains to make sure that higher order QCD corrections are not playing a dirty trick on us. If not, there will be one option left on the table....

13 comments:

Anonymous said...

How do you know that the asymmetry is largest for the dielectron events? I don't see the breakdown into electrons/muons in the public note.

Matti Pitkanen said...

For TGD inspired reckless speculations as a possible evidence for color excited Z boson for which the coupling to top pair has inherent parity breaking see this.

Luboš Motl said...

Thanks for your reports, Jester. This particular surprising story starts to look somewhat serious.

Jester said...

Anon, Figs. 42,43,44.

Jester said...

Oh, I see, they have taken down the link. On the CDF page there used to be a link to the note in which they give the break-down by flavors: A_ee = 0.270 ± 0.112, A_emu = 0.060 ± 0.077, A_mumu = 0.170 ± 0.102.

Matti Pitkanen said...

Second trial after censoring out of the first one: the reason cannot be related to content since also Lubos proposed after my posting also flavor changing neutral currents;-). It seems that situation is getting too hot to allow communication of TGD.


Flavour changing neural color and ew currents is one of the basic predictions of TGD and is due to the topological explanation of family replication phenomenon. This would be the first direct experimental evidence for TGD view about family replication (besides the dramatic indirect evidence provided by the success of p-adic mass calculations). See the blog posting and also the earlier posting.

Phil said...

Interesting stuff. A naive question from one who studies the skies... is it "just" a matter of labor to do the QCD calculation more accurately, or are there more fundamental technical difficulties to resolve first?

Jester said...

Not an expert here, but I suppose it's just the matter of labor. The asymmetry is currently known at NLO (alpha_s^3), it would be good to know the NNLO result.

Anonymous said...

@Phil and Jester

Yes, in principle it's "just" a matter of labor. However it's highly nontrivial to compute the NNLO QCD corrections to ttbar production. People are working on it, but it still might be a while (years more likely than month) before the full result is obtained.

On a sidenote, since the asymmetry is zero at leading order (LO), the NLO result for ttbar production just gives the LO asymmetry.
In addition, there are some results that use resummation and effective theory techniques to obtain parts of the higher order corrections to the asymmetry, but as far as fixed order perturbation theory is concerned, it's just leading order.

Kea said...

Ah, so flavor changing neutral currents are all the rage now, Lubos? But just a month or two ago you called me a crackpot for bringing up this subject ...

R said...

Offtopic answer to Kea. This has nothing to do with FCNC or your theories, but Lubos is not a phenomenologist. He doesn't know what is the rage in particle physics (defined as that part of high energy physics that has something to do with experiments). His string theory posts are serious, but the ones on LHC or phenomenology are often a bit naive and he frequently shows misunderstandings or overinterpretations of various results. For example, he seems to think that the Higgs mass cannot possibly be larger than some low number cited by LEP and Tevatron fits & searches.

Ulla said...

http://physics.aps.org/articles/v4/15

De Bunker said...

Oh no, not CDF and muons again. Ahem. Now we need to evaluate CDF's forward-backward asymmetry in charged pions, which certainly is not zero.