D0: 4 sigma like-sign dimuon anomaly!

About a year ago the D0 collaboration announced a surprising result. They compared the number of events with two positive muons and those with two negative muons. Once the contribution from kaons and pions decaying to muons within the detector is subtracted and some instrumental effects are taken into account, the number of positive and negative muon pairs is expected to be the same. Instead, D0 saw a 1% excess of events with 2 negative muons which represented a 3.2 sigma deviation from the Standard Model prediction. Yesterday D0 presented an update of that measurement based on 9fb-1, that is almost the full data set they have on tape. They obtain the asymmetry of −0.787% with an error of about 0.2%. The anomaly has grown to 3.9 sigma!.

The observed dimuon charge asymmetry is most likely due to asymmetric decays of B-mesons. Bottom quarks inside these mesons can decay as b → c μ- ν, and analogously an anti-bottom quark can decay to a positive muon. Most of the time the Tevatron produces pairs of bottom and anti-bottom quarks, each of them dressing into its own B-meson. However, neutral B-mesons can oscillate into its own antiparticles. If this happens, both original b-quarks may end up decaying to same-sign muons. Furthermore, if the oscillation probability violates CP, that is oscillating Bbar → B is more likely than the other way round, then the excess of negative muon pairs may show up. In fact, such an effect occurs within the Standard Model, but the predicted asymmetry is tiny, of order 0.01%. On the other hand, the asymmetry of the size observed by D0 requires new sources of CP violation beyond the Standard Model. Like what? Like Z', W' charged Higgs, KK gluons, or whatever; we would need more clues to guess the right answer.

An important new element in the latest D0 analysis is the study how the asymmetry depends on muon's impact parameter with respect to the primary vertex of the collision. Muons from B-meson decays often have large impact parameters because decay happens picoseconds after production. On the other hand, muons from kaon decays have typically small impact parameters because the mother kaon usually comes straight from the collision point. Thus, selecting events with large impact parameters enriches the sample with dimuons from B-mesons decays. D0 concludes that the dependence of the asymmetry on the muon impact parameter is consistent with the hypothesis that it indeed originates from B-meson decays, and not from some mundane background. Moreveor, the cut on the impact parameter also affects the relative fractions of Bd and Bs meson decays in the dimuon sample (these fractions are about 50-50 without the IP cut, but due to different oscillation parameters more Bd mesons spit muons with large IP). Thus one can put better constraints on separate contributions of Bs and Bd mesons to the asymmetry. The result is this plot:
The axes are the semileptonic decay asymmetries of the Bd and Bs mesons. The pink band is the fit to the observed dimuon asymmetry without the IP cut, while the ellipse takes into account the input from the IP measurements. Unfortunately, we still cannot tell whether the asymmetry is due to Bs mesons, or Bd mesons, or both, which is of primary importance for theoretical interpretations of the anomaly.


So have we discovered new physics yet? Alas, recent history teaches us not to celebrate before the signal is confirmed by an independent experimental group. The CDF collaboration had an anomaly even larger than 4 sigma which did not stop D0 from ruthlessly shooting it down. The rules of the wild west suggest that CDF may attempt the same with the D0 pet anomaly, after which they all meet at the O.K. Corral. But maybe this time it'll be different? Maybe this time it's for real? We may learn more later this year, either from CDF or from the LHC. Actually, the LHCb experiment promised to deliver a complementary evaluation of the B-meson decay asymmetries by measuring the B →D μ ν decay rates. Because of systematic effects they find it easier to determine the difference of the Bd and Bs meson semileptonic decay asymmetries (while the D0 dimuon asymmetry depends roughly on the sum thereof). With 1fb-1 of data their expected sensitivity corresponds to the thin gray band in the plot on the right. One more reason to bite our nails while waiting for the next LHC results!

Meanwhile at the LHC

The excitement about the CDF bump is subsiding so we can relax and look back at the LHC. A lot is going on there, although the best of the action will occur later this summer.

• As everyone knows, the luminosity collected by ATLAS and CMS just passed the 1 fb-1 benchmark which had originally been the goal for the entire 2011. The LHC management is running a policy of making careful projections which can later be spectacularly surpassed. We all agree it's better than the previous policy of bold projections and spectacular catastrophes.
• Unofficial predictions for the luminosity at the end of 2011 are Gaussian distributed with a peak around 5 fb-1. It means that, this year we'll probably learn whether the Standard Model Higgs exists or not! That's massive.
• Several results using a good chunk of this year's data, around 200 pb-1, have already emerged from ATLAS on the occasion of the PLHC conference in Perugia 2 weeks ago. The most interesting result is the t-tbar invariant mass spectrum. The most disappointing, too. If Tevatron's excess in the top quark forward-backward asymmetry is due to a 1-2 TeV heavy Kaluza-Klein gluon the t-tbar spectrum should display a peak or an excess at the tail. Alas, nothing there, as you can see.
• About the update on searches for Z' decaying to leptons, see Tommaso's blog. There's also a new search for W' in the muon + missing energy channel. In those cases also nothing, only vague rumors.
• Meanwhile SUSY searches are continuing at full steam. New analyses in the jets+met and jets+lepton+met channels are out. From the plot you can read that, for equal squark and gluino masses, the limit on the masses is above the magic threshold of 1 TeV. If the squarks are decoupled the limit on the gluino mass is slightly less stringent, about 750 GeV, and similalry the other way around. Children take about 6 years to realize Santa Claus does not exist, the LHC may be quicker than that.
• Theorists, on the other hand, are trying to understand the deeper meaning of the LHC limits on SUSY. The conclusion is that SUSY must be just behind the corner, just a little bit more, one last effort, and we'll see it. One should note that preference of the global fits for light superparticle masses is driven by one measurement: the long standing 3 sigma excess in the muon anomalous magnetic moment. Interestingly, a recent paper reevaluates the theoretical contributions to the muon g-2 and concludes there is no excess whatsoever. I am not in a position to judge whether the paper is correct, drop your comment if you are.
• Meanwhile, ATLAS and CMS keep posting papers on arXiv which use only the meager last year's harvest of 35pb-1. At this point it feels like offering ZX Spectrum in an Apple store.

The next big dump of LHC results is bound to happen for the EPS conference in Grenoble end of July. Expect numerous new physics analyses using 1.x inverse femtobarn of data. It'll rock!

D0: no bump

This result has been known to all blog readers since yesterday, now it's officially out. The D0 collaboration just released a new analysis of the dijet invariant mass spectrum in W+2j events using 4.3 fb-1 of data. Recall that CDF, looking at the same final state, saw an unexpected bump in the dijet spectrum near 150 GeV which might be a sign of new physics beyond the Standard Model. D0 closely follows the CDF analysis. Their conclusion: no bump.
They place the 95% CL limit on the cross section of a hypothetical 145 GeV dijet resonance at the level of 2 picobarns. On the other hand, CDF estimates that fitting their signal requires a larger cross section, of order 4 picobarns, which D0 excludes at 4 sigma. It's not straightforward to draw firm conclusions from these numbers. Although both experiments use almost identical kinematic cuts, they have different detector response, quality cuts, etc., which leads to different results. For example, D0 has about 30% more events than CDF near 150 GeV (after scaling with luminosity). It would be illuminating to run realistic models explaining the CDF bump through both detector simulations in order to compare the efficiencies. That would allow us to quantify the discrepancy between CDF in D0 (it may well be much smaller than 4 sigma). In any case, not seeing any excess in the D0 data puts a lot of strain on any new physics explanation of the CDF excess.

For the moment I don't see any catch in the D0 analysis. Previously, D0 employed differential reweighting to model the W+jets background which might have scaled away any excess; however, the present analysis avoids this step. Also, the D0 data match the Standard Model predictions much better than the CDF ones, also away from the 150 GeV region, which suggests they have backgrounds under better control. Looking closer at the plots, it seems the main difference between the 2 experiments is their estimate of the QCD multijet contribution (kudos to Jay for that observation), but at this point it is not clear if this can explain the CDF bump.

In conclusion, the 2 Tevatron experiments got into an epic standoff. One holds 4 aces in his hand. The other says it's a cheat. We need a shootout to decide who's right :-)

The story is of course all over blogs, see e.g. Tommaso, Michael, Georgios, Peter, Lubos, Sean for more coverage.

More details about the CDF bump

These days the next round of LHC results is beginning to emerge but for at least another week all eyes will be on the Tevatron. Recall that the CDF collaboration is observing an unexplained bumpy feature in the invariant mass spectrum of jet pairs in events with a W boson. Last Thursday CDF released an online note describing the latest update based on 7.3 fb-1 of data. The significance of the excess has increased to 4.1 sigma. The note describes a number of additional checks that the authors of the analysis have made to exclude background mismodeling as the origin of the bump. In particular, it seems that neither the standard model top quark, nor a difference of jet energy scales between quark and gluon jets can be responsible for the excess. Moreover, the excess persists (albeit a bit smaller) when a different Monte Carlo program is used to simulate the background. All in all, currently none of the known sources can explain the peak in a way consistent with all data. It must be a more subtle detector effect, or new physics.

Furthermore, the note presents a number of kinematic distributions of the events in the
window 115 < M_JJ < 175 GeV where the excess is the largest (thanks guys!). One plot makes you jump in your chair: It shows the invariant mass of the sum of the 4-vectors of the 2 jets, the lepton, and the neutrino, the latter reconstructed from the missing energy. If they all originate from one mother resonance, as suggested by a class of models, the invariant mass should reproduce that resonance mass. Indeed, the plot above shows a clear excess just below 300 GeV. This hints at another heavy particle being produced at the Tevatron, which then decays to a W boson and a 150 GeV particle who is directly responsible for the dijet bump. However, the plot on the right, which shows the same distribution but without subtracting the background, tells you that the standard model also peaks around 300 GeV (as a results of the imposed cuts) which makes the peak less trustworthy. If not for that, we would already be dancing on the streets and indulging in wild orgies.

It's also worth looking at another plot showing the transverse mass of the lepton + neutrino system. If the two come from a decay of a W boson, as tacitly assumed in the analysis, that distribution should have an endpoint at m_W = 80 GeV. Since most of the excess is below 80 GeV we can conclude that most of the times the 150 GeV resonance is accompanied by a genuine W boson. This excludes a more exotic class of models I mentioned where the lepton and the neutrino originate from the same particle that is responsible for the bump.

What's next? We are of course dying to hear the story from D0 and from the LHC experiments. The update from D0 is imminent. We expect it to be announced on June 10 at the Wine&Cheese seminar in Fermilab (as a general fact, wine facilitates communication between experiment and theory). As for the LHC, a blog post on Quantum Diaries points to an ATLAS note based on 33 pb-1 of data which roughly repeats the CDF analysis and finds no excess. However this means nothing: ATLAS simply had no right to see the hypothetical CDF bump in their 2010 data. First of all, the larger production cross section at the LHC (5 to 40 times, depending on the production mode) combined with the better efficiency does not make up for the 200 times smaller luminosity. Moreover, the W+jets background at the LHC is about 40 times larger. For these reasons you need a larger data set to make any conclusive statement. Suppose the CDF excess indeed originates from a 300 GeV mother resonance. If that resonance is produced by gluon-gluon collisions then the LHC should be able to see the excess already in 200 pb-1, which is the amount of data used in the most recent analyses. I'm sure hundreds of people are looking into this as we speak and as soon as the peak appears it will be pasted into Peter Woit's blog ;-) If, on the other hand, that resonance is produced in quark-antiquark collisions then we need to wait for 1 inverse femtobarn to see a significant excess at the LHC. In any case, the floor should be swept by the end of this summer, one way or another.

CDF: Wjj bump almost 5 sigma!!!

Today at the conference Rencontres de Blois the CDF collaboration presented an update on the invariant mass of 2 jets produced in association with a W boson. Recall that 2 months ago CDF posted a paper based of 4.3 fb-1 of data claiming that this observable displays an unexpected bump near 150 GeV with a significance of 3.2 sigma. The bump could have been a fluke, an accounted for systematic effect or surprising new physics. Now the first option is no longer on the table: the same bump is also present in the more recent data with a large statistical significance. With 7.3fb-1 of the Tevatron data, after subtracting the known Standard Model backgrounds other than the WW and WZ production, the distribution of the jet pair invariant mass looks like this:The peak has become more pronounced! CDF quotes the significance of 4.1 sigma (the number 4.8 sigma I quoted earlier takes into account only statistical uncertainties; after including systematic uncertainties the significance drops to 4.1 sigma). In a collider experiment, such a huge departure from a Standard Model prediction is happening for the first time in the human history :-) I don't have to stress how exciting it is. However we're not celebrating the demise of the Standard Model yet, not before an independent confirmation DZero or from the LHC. In any case, this summer is going to be hot.

For possible theoretical explanations of the bump, see my previous badly timed post. In the Blois slides CDF adds one important new piece of information. They say the bump cannot be due to the Standard Model top quark background, contrary to what was suggested in a couple of theory paper. Basically, there is no sign of enhanced b-jet content in the excess events, and in any case the top quark endpoint would show up below 150 GeV due to different jet energy scale corrections for b-jets.

Update: CDF has released more plots and the note describing the update.

Theorists vs. the CDF bump

Almost 2 months ago the CDF collaboration published their analysis of the events with exactly 2 jets, 1 lepton, and missing energy. These are vastly dominated by boring Standard Model processes where the W boson is produced together with jets and subsequently decays to an electron or a muon and a neutrino. A surprising feature showed up in the distribution of the invariant mass of the jet pairs. After subtracting the Standard Model background, CDF observed a bump near 150 GeV with a significance of 3.2 sigma. Obviously, theorists rushed to interpret the bump in term of physics beyond the standard model. The CDF result hints to a new particle with a mass of around 150 GeV, a significant coupling to the light quarks and a tiny coupling to leptons; the remaining details are left up to our imagination . Here is a selection of the educated guesses that appeared in about 50 papers to date.

The first thing that comes to mind is Z' - a new neutral gauge boson coupled to the left-handed quarks. This is a valid possibility provided Z' is leptophobic, that is to say, its coupling to electrons is less than about 0.05 to avoid constraints from the LEP experiment. There is some tension with the constraints from the UA2 experiment that was operating some 30 years before christ and made a search for a narrow Z' in the dijet channel. The UA2 limits on the Z'-quark coupling translate to a constraint on the W+Z' cross section at the Tevatron that allows one to explain only about 60 percent of the events observed by CDF. However, given the large uncertainties involved in the CDF measurement and in interpreting the UA2 results, the Z' option remains open. One should also note that nothing in the data tells us the new particle is a vector boson, it could just as well be a scalar.

To ease the UA2 constraints one can turn to another class of model. Quite generally, the Tevatron may produce a ≥ 250 GeV mother resonance who decays to a W boson and a 150 GeV daughter resonance. The latter subsequently decays to 2 jets who are observed by CDF. Several proposals for the mother and daughter exist: a technirho meson decaying to a technipion and a W in a version of technicolor, a sbottom decaying to a stop and a W in R-parity violating supersymmetry, a charged Higgs decaying to a neutral Higgs and a W in two-Higgs doublet models, a weak doublet color octet in the Manohar-Wise model, etc. The striking prediction of this class of models is that not only the invariant mass of the jets but also of the entire final state should display a resonance. CDF looked at the invariant mass of the 2 jets + lepton + missing energy vector and found it consistent with background only, but it is not clear if this excludes the presence of a mother resonance (the presence of the missing energy introduces larger systematic uncertainties than for the jet pair mass).

One can also imagine a more intricate class of models where the lepton and the missing energy in the CDF excess events come not from a usual W boson but from some other particle decaying to an electron and a neutrino. For example, this paper explains the excess by a production of a pair of supersymmetric winos of which one decays, via R-parity violation, to a charged lepton and a neutrino, and the other decays to 2 jets. This possibility may be excluded by analysis the distribution of the transverse mass of the lepton+missing energy subsystem.

Finally, one should mention those who are trying to spoil the party. From the very beginning many have cast doubts on the CDF analysis as it requires a perfect control over the overwhelming Standard Model backgrounds. One thing is that even the Standard Model W/Z peak in the observed jet mass spectrum, arising due to the well known contribution of the WW and WZ production processes, does not seem to be very well described by the simulations. Furthermore, by eye it seems that shifting the jet energy scale a few percent upwards, which would correspond to shifting the whole data curve to the right, allows one to get rid of the excess (the authors of the analysis reply that raising the jet energy scale makes additional events pass the analysis cuts, so that naive shifting of the curve is not correct; they say a 3 sigma excess persists even when the JES is scaled up by 7 percent). Another attempted explanation is that the apparent excess is in reality the Standard Model top quark. When a top quark decays hadronically, t → W b → jjb, the invariant mass of the 2 light jets of course peaks at the W boson mass of 80 GeV, however the invariant mass of the b-jet and one of the light jets has the distribution peaking near 150 GeV (the endpoint is Sqrt[mTop^2- mW^2] = 155 GeV), suspiciously close to the CDF bump. Thus, the excess may be due to the semileptonic t-tbar or single top production where one or more additional jets are missed at the detector, assuming the Monte Carlo simulations of that background have been (rather grossly) mismodeled.

So this is where we stand today. The situation may or may not be clarified when more data arrive. The updates from CDF and D0 are imminent. Someone will call a bluff? Or someone is holding an ace up his sleeve? Stay tuned for the next episode.

AMS is on

AMS-02 is up and running, and first events have already been twitted to the Earth. AMS is a full fledged particle detector attached to the ISS whose goal is to measure the cosmic ray spectra. The mission has been plagued by ill fate (delay due to the Columbia crash, scrapping of their superconducting magnet), now the road seems to be clear at last. The final preparations and the launch have been widely reported in the mainstream media, however my impression was that the actual science that AMS may accomplish was not clearly exposed. Here is my understanding of what AMS could teach us.

The official page of AMS lists the following scientific goals

• Search for primordial antimatter
• Search for dark matter
• Search for exotic forms of matter
• Study of the cosmic ray composition

The first point situates somewhere between Sam Ting's fixation and crackpottery. AMS will search for anti-helium nuclei arriving from the outer space. Unlike antiprotons, positrons and anti-deuterium, heavier anti-nuclei are not expected to be produced by cosmic ray collisions; anti-helium would have to be produced by astrophysical objects made of anti-matter. The problem is that we know there is no such thing: all the primordial antimatter annihilated with matter around 1 sec after the big bang. This view is not only the consequence of the current cosmological model, but it is also firmly supported by several independent observations, such as the cosmic gamma-ray spectrum, the cosmic microwave background, and the near perfect agreement between the predictions of nucleosynthesis and the composition of visible matter in the universe. Given the current body of evidence, AMS has a better chance for a 3rd degree encounter than for finding primordial anti-matter.

The situation with dark matter is more subtle. The PAMELA and FERMI satellites launched in the previous decade have been providing us with precise measurements of the high energy cosmic ray spectra. One thing we definitely have learnt is that it is painstaking to search for dark matter this way. Several excesses over theoretical predictions have been reported so far: PAMELA's positrons, Fermi's electrons, Fermi's photons from the galactic centre. They all have a plausible interpretation in terms of models of dark matter and an equally plausible interpretation in terms of boring astrophysical phenomena. AMS may provide more input regarding the high energy spectra. As can seen in the plots of the projected sensitivity, after 10 years of data taking they expect to extend the measurement of the positron and antiproton spectra up to almost TeV (compared to the current reach of PAMELA of about 200 GeV). It's hard to say if these projections are realistic, since it is not clear how much the resolution at high energies is degraded due to the replacement of the superconducting magnet by a weaker permanent one. Assuming they are realistic, particle physicists will be able to refine their models of dark matter, and astrophysicists to refine their models of pulsars. In any case, the chances for a smoking gun signal of dark matter appear slim at this point.

Nevertheless, there is one area where AMS is clearly superior to all previous experiments. The instrumentation of AMS includes a calorimeter, trackers, a Cherenkov detector and a time-of-flight detector to measure the energy, charge and mass of incoming particles. All this gives them very good particle identification, in particular they can easily separate heavier nuclei from much more numerous protons and helium nuclei. Flux ratios of various heavy nuclei, for example the boron-to-carbon ratio, are an important input for the models of cosmic ray production and propagation. Furthermore, if there exists exotic matter with distinct charge-to-mass ratio, for example the hypothetical strangelets with small Z/A, AMS is well equipped to identify it.

In summary, high energy astrophysics is a crowded field, and AMS is unlikely to turn it upside down. Their best shot for a spectacular discovery is exotic forms of matter with distinct Z/A ratio, provided they exist. Furthermore, if AMS and the ISS last long enough, and if the performance of the detector is as good as they promise, they should be able to extend PAMELA and FERMI measurements of the antiproton and positron spectra to higher energies, which may or may not clarify the origin of the positron and electron excess in PAMELA and Fermi. In the worst case AMS will sort out the spectra of heavier cosmic ray nuclei, providing valuable input for cosmic ray propagation model. Critics may complain that 2 billion dollars for tuning GALPROP is a lot. Optimists may stress that so far it's the only hope for returns from the 200 billion dollars sunk into the ISS.

Figures are taken from the talk of Andrei Kounine at TeVPA'10.