CDF Collab. at Fermilab: WW/WZ Prod. Lepton+Jet Decays

[Andy Resnick]

A student forwarded this to me:

http://arxiv.org/abs/1104.0699

This isn't my field- can someone help walk me through it? For example, to start:

"The CDF collaboration at the Fermilab Tevatron collider recently measured the cross section for the production of either an additional W or a Z boson in association to a W (WW or WZ diboson production) from a lepton plus jets final state described in Ref. [3]. One of the two methods described in that work uses the invariant mass of the two-jet system (Mjj) to extract a WW + WZ signal from data. Here we perform a statistical comparison of that spectrum with expectations by including additional data and further studying the Mjj distribution for masses higher than 100 GeV/c2, with minimal changes to the event selection with respect to the previous analysis. We find a statistically significant disagreement with current theoretical predictions."

What process is being probed by looking for a WW/WZ event? What is a 'jet' final state? What is the relevant difference between lepton+jet and two-jet decay modes?

The popular press is all gaga (and useless). TIA.

 

[negru]

It's just one of those 3 sigma results from the dying Tevatron. Nothing to see here

Jets come from gluons or quarks which hadronize (since they carry color and due to confinement can't be observed) and create lots of hadrons going in one direction.

Apparently this signal can mean either a new type of higgs, a new type of particle altogether, or even a new force. But if you do your statistics right, you'll know that 99% chance of it being correct translates to 99% of it being incorrect if your accelerator is getting killed soon.

 

[ZapperZ]

There will be a live video stream of a seminar presented by Viviana Cavaliere, University of Illinois at Urbana-Champaign, at 4:00 pm Central today, on this result.

http://vms-db-srv.fnal.gov/fmi/xsl/VMS_Site_2/000Return/video/r_livelogicindex.xsl?&-recid=573&-find=

Zz.

 

[Haelfix]

I'm not so sure, this one actually has a few of us interested and comes as a bit of a surprise. This process and its backgrounds are relatively well understood, so I find it unlikely we made a significant background estimation error, especially at Tevatron.

Still spikes come and go in particle physics.

 

[fzero]

Tommaso Dorigo always gives a nice summary of interesting experimental particle results on his blog. In this case, since he is part of CDF, he's particularly familiar with the result and has already released an entry at: http://www.science20.com/quantum_diaries_survivor/new_massive_particle_some_kind_higgs-77857 He says that he will update that posting after the above seminar takes place.

 

[kloptok]

...

What process is being probed by looking for a WW/WZ event? What is a 'jet' final state? What is the relevant difference between lepton+jet and two-jet decay modes?

...

The popular press is all gaga (and useless). TIA.

(I don't know about your background so you'll have to excuse me if this is piece of cake for you!)

A W- can decay into a lepton and the corresponding antineutrino (or the opposite if we're talking about W+), and it can also decay into a quark-antiquark pair (where the charge of the pair sums up to -1 or +1, for example a W- can decay into a d+ubar pair). The Z can decay into lepton-antilepton (electron-positron or similarly) or quark-antiquark (where the quarks must be of the same flavor).

The quarks can't appear in isolation and so they immediately hadronize creating 'jets', containing a large number of color-neutral, reasonably long-lived hadrons emitted in a collimated cone with respect to the initial quark.

As to your questions, I would assume that for a WW or WZ pair, a 'jet final state' is a final state where both the W:s (alternatively the W and the Z) have decayed into quark-antiquark pairs, resulting in 4 jets, 2 for each W boson. A jet+lepton decay mode is when one of the W:s have decayed into a quark-antiquark pair, and the other into a lepton-antilepton pair (alternatively the W decays into quark-antiquark and the Z to lepton-antilepton or vice versa).

I believe that final states containing leptons are good experimentally since it is easier to measure the energy of leptons. Also, the experimental background is smaller, since there are a lot of decays that lead to jet final states, creating a large background for the jet final states. Furthermore, you would immediately know which two jets belong together (coming from the decay of one of the W or Z), since there are only two jets. In the four-jet final state you have to try your way with different combinations to pair the jets correctly.

I can't help you concerning what process the CDF collaboration is investigating through this, but I guess it's about the Higgs as usual.

 

[daschaich]

... I guess it's about the Higgs as usual.

Well, Viviana was careful not to make any claims in her talk this afternoon, but the preprint linked in the original post argues (on page 7) that this is not the Higgs (at least not the standard model "Higgs as usual"), both because the cross section is too large, and because there aren't enough b-jets. Viviana's thesis also mentions (page 151 of the pdf) that the observed cross section times branching ratio is about two orders of magnitude larger than that of the standard model Higgs.

When I saw Estia Eichten, Ken Lane and Adam Martin mentioned in the talk, I started wondering whether this could be some sort of technipion signal. Ken's an eternal optimist who likes to point out that exclusion limits on technipions have been stuck around 125 GeV for the last five years or so (at least since the DZero study hep-ex/0612013 with 388 inverse picobarns).

However, it sounds like the basic technicolor straw man model isn't a viable explanation, either. A couple of the questions/comments from the audience after the talk mentioned that a technicolor-related model attempting to explain this bump as a new particle will appear on the arXiv this evening, but had to become pretty non-standard to work. Viviana mentions both technipions and MSSM Higgs bosons in her thesis, as possibilities.

So far the most worked-out model addressing this bump that I've seen is 1103.6035, which proposes a light Z-prime coupling dominantly to up and down quarks. I haven't looked at it very closely (I know one of the authors, who suggested I watch Viviana's talk), but it's notable in that it tries to explain three of the recent 2--3sigma Tevatron observations (top quark forward-backward asymmetry, and excesses of three or more b-jets, in addition to this W+jets bump).

All in all, I'm not excited by this result, though I do find it interesting. I'm not sure I buy CDF's justification for submitting a 4.3 inverse-femtobarn PRL while rerunning the analysis with the 7 inverse femtobarns currently available, though. Viviana argued that the additional data wouldn't be able to kick the signal up to 5sigma, but it seems to me that the more pertinent question is whether or not the additional data will knock it down below 3sigma (which I expect). Nobody at Fermilab could still be hoping for more Tevatron running, could they?

 

[daschaich]

...A couple of the questions/comments from the audience after the talk mentioned that a technicolor-related model attempting to explain this bump as a new particle will appear on the arXiv this evening, but had to become pretty non-standard to work...

Well, here it is. It looks more like the usual technicolor straw man model than I was expecting based on the comments during the talk. The only complication seems to be the invocation of a CKM-like mechanism to suppress technipion decays to b-jets, which reduces the signal cross section to about an order of magnitude below that measured (still an order of magnitude above the standard model Higgs, and with much more wiggle room thanks to the strong dynamics).

OK then... "eternal optimist" indeed. I even know where Ken is looking for the technivectors decaying into the W+technipion in this scheme: both CDF and DZero recently reported less-significant bumps in dielectron spectra around 300 GeV (1008.2023 and 1103.4650), which Ken has been pointing out to anyone he can corner. (I hear, though, that an ATLAS paper without a 300 GeV bump, based on 2010 LHC data, is making its way through the approval process.)

 

[kloptok]

Interesting to hear about this from more experienced people than myself.

Would an explanation of this demand non-SM physics? It sure seems so, since all explanations mentioned so far have been about Technicolor, new particles etc. Could a bump in the invariant mass plot be something else than a particle decaying? If so, what?

Also, are there results from LHC supporting this data? dashaich mentioned something about an ATLAS paper due soon, 'without a 300 GeV bump'. What does this 'without a 300 GeV bump' mean?

 

[Lapidus]

here is an NY Times article on this
http://www.nytimes.com/2011/04/06/science/06particle.html

“If it is real, it would be the most significant discovery in physics in half a century.”

 

[daschaich]

Would an explanation of this demand non-SM physics? It sure seems so, since all explanations mentioned so far have been about Technicolor, new particles etc. Could a bump in the invariant mass plot be something else than a particle decaying? If so, what?

There are (at least) two reasonable explanations other than a previously-unobserved particle. First, it could be a statistical fluctuation that will disappear when CDF adds more data to the analysis (or DZero performs its own analysis). It's statistically unlikely for 3sigma signals to be fluctuations, but not unheard of.

Second (and more interesting), it could be a sign that other "background" processes producing the same final state signals are being incorrectly modeled around this energy (where this modeling includes accounting for how the final states interact with the detector, and the reconstruction of events from the resulting detector data). I agree with Haelfix that the Tevatron has been operating long enough and this analysis is straightforward enough that such mismodeling seems unlikely -- which is why it would be more interesting (at least to those in the biz).

However, it's worth emphasizing that this analysis is looking for a relatively small difference between two relatively large numbers (the observed and expected number of events). Even a fairly minor mismodeling may be able to wipe it out, and this also has implications for the LHC. Because (1) the expected standard model backgrounds (apart from the Higgs) have been accounted for in the modeling and (2) the remaining signal doesn't behave like the standard model Higgs, non-SM physics would be the more likely (though still not guaranteed) explanation if the signal holds up to additional data and scrutiny.

Also, are there results from LHC supporting this data? dashaich mentioned something about an ATLAS paper due soon, 'without a 300 GeV bump'. What does this 'without a 300 GeV bump' mean?

The LHC is probably going to need a lot more data before it can say anything about this signal, depending on what it is (if anything). There are two main differences between the Tevatron and LHC. First, the Tevatron collides protons with antiprotons, while the LHC collides protons with protons. If CDF is seeing a new particle, its production could be suppressed or enhanced at the LHC relative to the Tevatron, depending on how it is produced and how it couples to other particles (particularly quarks vs. gluons).

Second, the LHC operates at a significantly higher energy (currently more than three times the Tevatron energy). To a first approximation, the additional energy available should increase W+jets backgrounds in the relatively low-energy region of 120--160 GeV, making it harder to pick out (or exclude) a small excess. A relatively low-mass standard model Higgs boson with mass around 120 GeV is going to be harder for the LHC experiments to discover than a SM Higgs boson with larger mass. Since the LHC experiments hope to have enough data by the end of 2012 to explore the entire range of potential Higgs boson mass, they may be able to say something about this signal by that time as well.

In early LHC running, it will be much easier to see simple and clear signals at higher energies than the Tevatron is able to reach. The two-electron signals that show the (statistically insignificant) excess around 300 GeV at the Tevatron are much easier to study at the LHC -- apparently to the extent that ATLAS can improve on CDF/DZero results with less than one percent as much data. (Reviewing my notes, however, I noticed that the ATLAS study I heard a talk about is actually for two-muon signals, rather than two electrons. Something seen in one should be seen in the other, though -- if it is really there.)

 

[zaybu]

Technicolor at Fermilab

New findings at fermilab suggests that technicolored particles were discovered.
http://arxiv.org/PS_cache/arxiv/pdf/1104/1104.0976v1.pdf

Does anyone know anything about this theory?

 

[daschaich]

...1104.0976v1.pdf

Does anyone know anything about this theory?

That's the one I discussed in post 8 in this thread. I described it as "eternal optimism", but I would also accept "castle made of sand".

I'm reasonably familiar with this technicolor straw man model, which more or less dates back to http://dx.doi.org/10.1016/0370-2693(89)91265-3 in 1989 (though the references in 1104.0976 would be better places to check out this phenomenology). This CDF result looks reasonably consistent with what I would expect from this model (up to the b-jet suppression I mentioned last night), but claiming "that technicolored particles were discovered" is extremely premature (and is not the language used by the preprint, which "proposes" a "hypothesis", even though they argue it is "the most plausible new-physics explanation").

I guess it's sort of obvious to suppose this bump could be a technipion (after all, it occurred to me during Viviana's talk), but I worry that if the bump goes away, too many papers like 1104.0976 may give Ken a reputation of "crying wolf". Of course, if it turns out to be right, he can start looking into hotels in Stockholm, thanks to the 30+ years he and Eichten have been working on technicolor and its collider phenomenology. Well, I get a citation out it either way.

 

[negru]

Well the CDF already has a reputation of crying wolf, no one here at least was marginally interested in the result...at most the D0 people were upset because now they have to waste time rechecking data every time CDF cries wolf..

 

[daschaich]

Well the CDF already has a reputation of crying wolf, no one here at least was marginally interested in the result...at most the D0 people were upset because now they have to waste time rechecking data every time CDF cries wolf..

Out of curiosity, where is "here" and what other CDF analyses do you have in mind?

 

[negru]

Weren't there like 2 major announcements just in January, 2 or 3 sigma stuff? At least one of which the d0 analysis wiped out. And like you said, why wouldn't they include the extra data they have? She didn't even mention the possibility of the signal going away, when that was obviously what everyone was thinking about.

I'd also not giveway the location of the complaining d0 folks :) ..what I heard was from a grad student so I am already saying too much.

 

[negru]

Can anyone comment on the shift issue thing people are talking about on the usual blogs?

 

[daschaich]

Can anyone comment on the shift issue thing people are talking about on the usual blogs?

I don't monitor the usual blogs, but I assume this is the same issue as a question (the first question?) after Viviana's talk. Consider the upper-left panel of Fig. 1 in 1104.0699: in this plot of the unsubtracted energy distribution, the excess is visible above a rapidly-falling background spectrum. If there were some systematic effect causing the measured dijet invariant mass to be overestimated (that is, if some of the data should really be shifted one bin to the left), then the data would match the backgrounds perfectly, with absolutely no signal of anything else.

Viviana was very emphatic that this possibility has been studied six ways from Sunday (though that's my own phrasing), and is not a realistic explanation. It's such an obvious issue that I'm sure it was very carefully considered during CDF's internal review, part of the reason this analysis took something like 18 months to complete and get through approval. (That was Viviana's explanation for why the 4.3 inverse femtobarn data sample was used.)

This is a generic issue when looking for signals on the "shoulders" of large peaks, and part of the reason those analyses are difficult and need a lot of data (to say nothing of good energy resolution and understanding of the detector).

 

[negru]

Ok this might be a dumb question, but once you have all your methods fixed, shouldn't adding in more data be trivial?

 

[daschaich]

Ok this might be a dumb question, but once you have all your methods fixed, shouldn't adding in more data be trivial?

I would expect it to be a matter of weeks rather than months, but this is getting beyond my area of expertise. You might want to ask your DZero friends (and let us know what they have to say).

 

[Andy Resnick]

Tommaso Dorigo always gives a nice summary of interesting experimental particle results on his blog. In this case, since he is part of CDF, he's particularly familiar with the result and has already released an entry at: http://www.science20.com/quantum_diaries_survivor/new_massive_particle_some_kind_higgs-77857 He says that he will update that posting after the above seminar takes place.

That's remarkably helpful- thanks!

 

[tom.stoer]

has been moved from the "beyond forum" to "particle physics"; let's see if this was a wise step :-)

 

[zaybu]

What I'm looking is if the data holds, is there any truth that technicolor is the next thing to study as some have claimed? I want to know before I start investing studying that theory.

 

[tom.stoer]

Has string theory to say anything regarding technicolor?

 

[Andy Resnick]

Tommaso Dorigo always gives a nice summary of interesting experimental particle results on his blog. In this case, since he is part of CDF, he's particularly familiar with the result and has already released an entry at: http://www.science20.com/quantum_diaries_survivor/new_massive_particle_some_kind_higgs-77857 He says that he will update that posting after the above seminar takes place.

One thing that struck me from his blog entries is that the discrepancy between the measured data and the simulation could be due to two effects- 1) there is new physics (the Z'), or 2) the simulation has systematic bias (rescaling the jet energy).

How then can a rational distinction be made between the two options?

 

[erkokite]

Has string theory to say anything regarding technicolor?

Randall Sundrum models are AdS/CFT dual to some technicolor models.

 

[daschaich]

What I'm looking is if the data holds, is there any truth that technicolor is the next thing to study as some have claimed? I want to know before I start investing studying that theory.

Well, I feel that it's good to have at least qualitative familiarity with technicolor... and with the MSSM, and with the Little Higgs approach, and with any other more-or-less viable weak-scale physics beyond the standard model. I personally find the Z-prime model discussed above more interesting (though also more contrived), because it's less familiar to me and manages to push multiple 3sigma buttons.

I keep a close eye on the Wikipedia article on technicolor, which may be a quick way to get a basic introduction to the qualitative ideas behind the approach. Importantly, it also includes many good references: In general, Wikipedia is a good place to start learning about something, but a terrible place to finish learning.

Has string theory to say anything regarding technicolor?

Randall Sundrum models are AdS/CFT dual to some technicolor models.

In my experience, those tend to be the least-interesting sorts of technicolor models, where phenomenological constraints are particularly severe thanks to large N (unless you do some more-or-less ad hoc fiddling a la Hirn and Sanz).

In terms of string theory proper, I think of it living so many orders of magnitude above the electroweak scale where technicolor resides that I don't expect any directly relevant information from it for the foreseeable future.

...1) there is new physics (the Z'), or 2) the simulation has systematic bias (rescaling the jet energy).

How then can a rational distinction be made between the two options?

Well, that's the question (assuming it's not a statistical fluctuation, which is still what I expect). At present, I think the best we can do is consider what further information we would need in order to make such a rational distinction. For instance, independent information from DZero (or better yet the LHC) will be of great interest, whether or not a similar signal is seen.

Any models attempting to explain this signal also make other testable predictions (for the Z-prime these include the other 2--3sigma signals already mentioned, and the technicolor paper 1104.0976 also "suggest companion searches" to check the possibility of this new physics scenario). That's an obvious way to explore the possibility of new physics, and if any of the predictions were confirmed experimentally it would encourage a rational decision that there is new physics at play.

Similarly from the other direction, I'm sure folks are carefully considering possible sources of systematic effects that have not yet been accounted for. Simply rescaling the jet energy does not look like a viable explanation, but it's possible some of the backgrounds have been mis-modeled, which would be interesting in its own right (if not to the NY Times).

 

[tom.stoer]

Randall Sundrum models are AdS/CFT dual to some technicolor models.

Any references?
Is Randall-Sundrum - strictly speaking - string theory, or string-theory inspired?

 

[daschaich]

Any references?
Is Randall-Sundrum - strictly speaking - string theory, or string-theory inspired?

I would suggest the Hirn and Sanz "Holographic Technicolor" paper linked in my previous post, and more recent articles citing it. I consider AdS/CFT string-theory inspired, but others' attitudes may vary.

 

[piareround]

So I take it from the reading previous posts that we are not seeing gravity finally seperating into two forces, and thus more than one particle, as Hubble's Law predicted should eventually happen to our galaxy in the next billion years. Also I assume this probably isn't Dark Energy either.

Is this correct?

(Sorry I know its a stupid question, but other people have been asking questions of if this new particle is related Dark Energy in other forums. I am linking them to this thread to clarify, so if you guys could please be patient and answer honestly it really help. Thanks ^^; )

 

[Haelfix]

I don't buy the energy scaling argument that Tomasso mentioned. That's literally the first thing they look for, and while the correlation is impressive (nearly a perfect fit) I don't believe that this would have ever made it out of committee if that was the problem.

Thats not to say that it isn't ultimately the explanation, just that it wouldn't be so trivial.

 

[typical guy]

Does this 3 Sigma event match any of Lisi's new particles predicted in E8?

 

[Vanadium 50]

How then can a rational distinction be made between the two options?

You can't with this data.

I don't buy the energy scaling argument that Tomasso mentioned. That's literally the first thing they look for, and while the correlation is impressive (nearly a perfect fit) I don't believe that this would have ever made it out of committee if that was the problem.

As always, Tomasso never fails to disappoint. If he felt this was the explanation, and that the paper was wrong, he should have refused to sign the paper.

It's unlikely that the jet energy scale is off by several percent: there are too many other places it would show up. However, I have no trouble believing that the Monte Carlo is off by a few percent in the predictions.

Remember, the significance is not the probability that the peak is not a particle. It's the fraction of time the data would be at least this discrepant from the Monte Carlo if the Monte Carlo were perfectly correct.

 

[daschaich]

So I take it from the reading previous posts that we are not seeing gravity finally seperating into two forces, and thus more than one particle, as Hubble's Law predicted should eventually happen to our galaxy in the next billion years. Also I assume this probably isn't Dark Energy either.

Is this correct?

This signal is unrelated to gravity or dark energy.

Does this 3 Sigma event match any of Lisi's new particles predicted in E8?

No.

 

[piareround]

Thanks for reply that what I thought. :3