Another week, another inverse femtobarn

The Large Hadron Collider has settled into a steady period of running and is now delivering about one inverse femtobarn each week (actually it was about 0.98/fb in the last seven days including a single fill of over 200/pb yesterday.) There are 17 days until the cut-off for ICHEP (contribution list now filling out) , the biggest international particle physics conference of the HEP calendar that is being held in Melbourne from 4th July. If there are no snags they should still be able to reach the target of 5/fb delivered and the experiments will show new Higgs results with at least 4/fb at 8 TeV. That is about as potent as last years 5/fb at 7 TeV. If they get similar bumps at the same places as last year that should settle the existence of the Higgs at least unofficially, but unlucky statistical fluctuations could still leave the result hanging. Official discovery will probably have to wait a little longer. I think they will hold back from combining the new data with last years until both experiments can claim discovery. ATLAS and CMS have shown very similar sensitivity and resolution especially in the crucial diphoton channel despite completely different detector technologies. It would not be fair if one of them got the discovery first due to lucky statistics. They can conveniently time their combinations to avoid that hapening.

According to Paul Collier (commenting at LHCportal) they still have some scope to slowly increase the bunch intensity up to 160 billion protons per bunch for another 20% increase in luminosity. That is the limit for this year and realistically I think that 1/fb per week will be about the average amount delivered over the remaining 17 week of proton physics minus any time for recovery from technical stops, extra MD or special runs for TOTEM. They should comfortably reach their target of 15/fb total delivered to each of ATLAS and CMS.

Full Higgs combinations of ATLAS+CMS have not been seen since autumn last year when they had just 2/fb. Although they have not given any explanation for this there are several factors that come into play. The sheer complexity and quantity of all the data means that the big combinations require enormous amounts of computer resources (when done exactly). By time they could complete the calculation the experiments have usually added something new making the answer obsolete before it is ready. I imagine they can’t afford to waste their computer resources or manpower in that way with so many other things to be done. Another problem is that the two experiments have seen maximum peaks at slightly different Higgs masses. The favorite theory for this is that the energy calibration has worked out slightly differently through systematic errors in the two experiments. The consequence is that the combination only gives a small improvement in significance over the individual results. But I think the most important reason for not doing the full combination now is that they can reach discovery level this year with both experiments separately. It will make a more spectacular and convincing presentation of the discovery if they can do it that way rather than with a cross-experiment combination. I could be wrong but I think there is a good chance that they will do full combinations of ATLAS+CMS only during the long shutdown as part of a more detailed analysis to compare observations with the standard model predictions. Until then people will have to be content with unofficial approximate versions.

While you are waiting for ICHEP do not miss the meeting at Blois. The first session has some particularly interesting talks not least of which is a presentation by Nima Arkani-Hamed on why he thinks a 125 GeV Higgs smells like SUSY. It’s a 30 minute talk but there is already a YouTube version online where he crams it into 1 hour 42 minutes and still skips slides at the end. You are recommended to set aside the time to watch it.

Update 25-May-2012: They have now passed the milestone of delivering 1/fb in one week to ATLAS and CMS. Here is a table of the fills that did it with the latest still running.

fill start date time dur ATLAS /nb CMS /nb
2663 25.05.2012  03:30:00 06:00+ 100000+ 100000+
2660 24.05.2012  17:34:00 01:14 24315.4 24738.1
2658 24.05.2012  03:53:00 01:22 26748.5 27252.6
2657 23.05.2012  18:41:00 06:03 92480.8 93457
2653 23.05.2012  08:55:00 00:37 11738.2 11521.5
2651 22.05.2012  08:48:00 19:35 209131.5 201505
2649 21.05.2012  13:41:00 07:19 109045.9 111447.2
2648 20.05.2012  18:15:00 05:34 78680.4 80486.8
2646 19.05.2012  13:14:00 16:06 187771.5 194043.1
2645 19.05.2012  10:23:00 00:30 9981.2 10042.3
2644 18.05.2012  13:55:00 14:45 170798.8 176034.6
total 1020692.2 1030528

59 Responses to Another week, another inverse femtobarn

  1. publicfunding says:

    125 GeV Higgs smells like SUSY because otherwise many physicists will loose their comfortable and well paied positions. SUSY the last immortal, more immortal than the philosophical stone!

    • Philip Gibbs says:

      Many physicists in well-paid positions? Are you talking about Wall Street now because otherwise I am confused?

      • Lubos Motl says:

        I just got a nicely looking fresh preprint from a physicist – about a way to look at spinors in 4D and 5D – who is a derivative trader in London. 😉

  2. Tony Smith says:

    Elwell comment 24-05-2012 13:53:53 says:

    “Next fill with new injection scheme
    No more bunches colliding only in IP8”

    Why is that being done ?
    Is there some problem with IP8 and if so what is it ?


    • Philip Gibbs says:

      IP8 is ALICE. I think it is just that they dont really need more proton events now. By cutting them out they can make every bunch collide in ATLAS and CMS giving a few percent extra luminosity. ALICE will have its time at the end of the year with heavy ions.

    • Philip Gibbs says:

      Sorry my bad, IP8 is LHCb, they already cut out ALICE earlier, now they are cutting out some bunches that were only colliding in LHCb but this seems to be for technical stability reasons rather than to give the slight extra luminosity in ATLAS and CMS.

      • dd says:


        There were two reasons for the change.

        The first reason they removed IP8-only and IP5-only bunches is that the losses incurred some of the bunches were much higher the average (when monitored by the FBCT in bunch-by-bunch mode). These losses themselves aren’t problematic, but the BPMs (beam position monitors) were losing track of these bunches (because their bunch-by-bunch intensity was much lower) which caused a few beam dumps. It’s not easy (or perhaps not possible within MP paramaters) to change the calibration for the BPMs to account for this.

        The second reason is that LHCb is already on continuous lumi leveling, and there is always overhead between the max. inst. lumi in LHCb compared to their desired levelling. Instead of wasting the bunches to this overhead, they can be used to slightly increase the lumi of IP1/5.


  3. carla says:

    Do you know what’s the increase in production rate for the Higgs for the 15% increase in energy from 7 to 8Tev?

  4. ondra says:

    carla, its about 10 percent
    Tony, there seem to be problems with sudden intensity losses of those “private” bunches leading to beam dumps

  5. Tony Smith says:

    What would be some likely causes of intensity losses of IP8 private bunches ?


    • dd says:


      See my other response above. It’s not the losses that are the problem, it’s that the BPMs lose track of bunches when they experience higher losses than the majority of the bunches.

      In other words, the BPM system reports that a particular bucket has (effectively) no bunches in it. This is a serious problem from a machine protection point of view, so an automatic dump is issued.

      Rather than adjust the off-scale-low sensitivity of the BPMs (which may be difficult to impossible), they are just avoiding the situation by avoiding IP2/8-only collisions.

    • Philip Gibbs says:

      dd, thanks for your answers which are much more detailed and convincing than mine.

    • Kasuha says:

      LHCb collision point is slightly off compared to other IPs – it’s not in the middle of the cave but a bit to its side, I think by about 20 meters. Bunches which collide in both Atlas and CMS would be able to collide in LHCb only if its IP was in the middle of the cavern and vice versa – bunches which collide in LHCb with some of bunches colliding in Atlas and CMS cannot collide anywhere else because in Atlas, CMS and Alice they pass their counterparts 20 m from the IP. Bunches colliding only in LHCb have most of such ‘close encounters’ and may suffer strongest beam-beam effects.

  6. kevin says:

    Many people here seems to think that adding a sigma to 4 sigmas give 5 sigma but it is not the case. If you combine a 1 sigma positive result with a 4 sigmas positive result, you obtain roughly a 4 sigma result. A similar results holds for sensitivity. If an experiment is expecting a 4 sigma signal on the average, a so called fluke, which is a 1 or at worst a 2 sigma deviation from the expectation, then you mesure a 4 sigma signal. Because if you have bad luck, a 2 sigma fluke down substracted to a 4 sigma signal, you obtain maybe a 3.9 sigmas signal. So there will not be any fluke problem for experiments.
    The problem is real when you have a 2 sigma negative fluke over a 2.5 sigma signal expectation, which is the case now for individual channels with 5 fb-1 of data.
    I think that if both Atlas and CMS see 3.5 sigmas signals in ZZ and gamma gamma channels by combining there 2011 and 2012 data, the discovery will be claimed in July. That would give 4 signals with 3.5 sigmas significance from 2 independant experiments. If there are “2 sigma down flukes” then we’ll maybe have to wait for the end of the year.

  7. Ervin Goldfain says:

    How can LHC claim a “discovery” of the Higgs boson without overwhelming evidence that the signal represents a Higgs indeed and not something else? According to Prof. Strassler, Phase 2 of Higgs searches will have to validate production and decay rates (among other attributes) and is slated to begin in 2013:

    • Philip Gibbs says:

      There is really no big deal here. Obviously they will word the discovery carefully along the lines of “we have discovered a resonance consistent with the predictions for a standard model Higgs boson” then they will quantify and illustarate what they mean by that with numerous plots. Later papers will improve the agreement or perhaps show deviations more consistent with other models if they are really lucky.

      • Ervin Goldfain says:

        Of course they are going to be extra cautious with their statements. But that’s not my point. My point is that there will be no definitive conclusion on the discovery prior to finishing up all required cross-checks. And this might take us well past 2012.

      • Lubos Motl says:

        I’ve considered Matt an excellent phenomenologist – and, in fact, some kind of near-experimenter as well – for years. This makes it even more bizarre to notice that your description of these basic issues makes much more sense than his and, in fact, actively clarifies some confusion that inevitably follows from Matt’s comments that were misleading in effect if not in intent.

      • Lubos Motl says:

        Dear Ervin, I can’t believe that you didn’t understand Phil’s crisp and clear explanation. The process has several stages. Around Summer 2012, the experiments will have enough data to claim a 5-sigma deviation from the “Higgsless Standard Model” so they will have discovered *something*. And it will look as something very similar to – and with the available data, indistinguishable from – a Higgs boson that is needed, too.

        One obviously can’t wait with the discovery announcement up to the moment when all the behavior of this new particle is measured with a near perfect accuracy. In fact, this will clearly *never* occur. And in fact, it’s pretty much guaranteed that the interactions and decays of this new particle are *not* exactly equal to those in the Standard Model because the world isn’t exactly described by the Standard Model.

        But finding deviations from the Higgs in the Standard Model is another stage of the research which doesn’t affect the fact that by the end of the summer, the LHC detectors will have discovered a new particle that seems compatible with the Higgs boson. That means the SM Higgs or a Higgs in an extension of the Standard Model – which would be even more exciting – or a completely different new particle – which would be even more exciting than that. But given the absence of any other statistical evidence, the discovery of the 5-sigma deviation will hint at the SM Higgs which will obviously become the new null hypothesis to describe what’s happening with this resonance.

        New research and more accurate measurements will be needed to detect deviations from the SM Higgs hypothesis. That’s totally analogous to the fact that since the 1980s, we have called the 80 GeV and 90 GeV spin-one particles “W-bosons” and “Z-bosons” of the Standard Model even though we didn’t have a full proof of the Standard Model and we were, in fact, totally ignorant about the mass of its most novel particle that makes W-bosons and Z-bosons massive. But everything we could observe was compatible with the SM W-bosons and Z-bosons which is why we were calling these particles W-bosons and Z-bosons and we implicitly or explicitly said that they were particles of the Standard Model.


  8. Ervin Goldfain says:

    Dear Lubos and Phil,

    It seems that there is a wide difference in opinions and expectations on the imminent “discovery” of the Higgs boson this summer. You may consider dedicating a blog posting to discuss these issues.


    • Lubos Motl says:

      Dear Ervin,

      apologies, I don’t find it an excessively interesting topic for a new text so this comment is the replacement.

      In my optics, the Higgs boson has already been de facto discovered. The official discovery will take place when an experiment accumulates 5 sigma. If they decide not to join forces, this will only happen at the end of the summer and sometime in the fall,, the detectors will announce the official discovery of a resonance that is compatible with the SM Higgs boson, or whatever the wording will be.

      The alternative that the bumps we have seen are flukes seem utterly unrealistic to me. Of course, with some low probability that may be around 1 in 100,000, it may happen and the 125 GeV bumps may go away. In that case, there would probably have to be a different Higgs, e.g. one above 600 GeV or some strange composite one. When I add both theoretical and experimental discouragement for this scenario, the odds are one in a billion or trillion. I don’t think it’s too reasonable to spend much energy with such insanely unlikely things. And even if this scenario materialized, there wouldn’t be much to write immediately. We would have to wait for the discovery of the right thing that breaks the EW symmetry.

      I don’t really believe that there is a disagreement between physicists about the Higgs’ being de facto discovered. People I know agree that the new default assumption is that it’s been found – randomly: David Gross, Nima Arkani-Hamed, Gordon Kane, and many many others. Those who disagree don’t really have an opinion; I think it’s more accurate to say that they have an agenda. They either want to make the ongoing extra research of the “Higgs discovering” look more important than it is because they’re still a part of this business; or they generally want to claim that any question is “too close to call” even though the odds are above 100,000 to 1. It’s about their very different personality or whatever. But in science, what matters is the validity of the answers and be sure that I, Phil, and others will be shown right while those “too close to call” people will be shown as having missed the opportunity of making the right “prediction” or, more precisely, revealing a sensible evaluation of the past observations.


      • Giorgio says:

        The official CERN statement is that the probability of a statistical fluke is not 1 in 100 000, but up to 30%. See the full CERN explanation at (and others).

      • Philip Gibbs says:

        That paper is for ATLAS on its own with a very generous LEE. We are talking about unofficial conbinations of results from CMS and ATLAS and even Tevatron.

      • Giorgio says:

        The text says that CMS has similar results to ATLAS (which has 30% fluke probability). Whoever says that in combination, this yields 1 to 100 000 (4.4 sigma) is spreading nonsense or false information. The situation might be different in the summer 2012 – but so far it isn’t.

      • Philip Gibbs says:

        It would be a serious error to apply LEE to each experiment individually before combining which is what you are doing. In any case I regard the LEE used as over cautious. They should take into account that most of the mass range is excluded and LEE has to be caclulated on the basis of the range left. After combination that range is very small and LEE is almost eliminated.

    • nameab says:

      Have you heard the feasibility of Bohr orbit quantization for multi-electron,would you tell me your website?

    • Philip Gibbs says:

      I agree 100% with Lubos and just add this. CERN are setting themselves a very high standard of discovery here. In the recent talk Peter Higgs gave at Bristol he mentioned a comment that someone had made to him. They said that the significance of the present Higgs evidence is better than that for the last Nobel prize.

      They are setting a high standard firstly because they can, but also because nobody if infallible (pentoquarks, OPERA, etc). It is also sadly true that the world is full of Wormtongues who will tell politicians what they want to hear so that they cut funding. Just look at all the people who claimed the LHC would destroy the world. CERN must counter this by working to the high possible standards. If they relied on a combination some people would ask “how do we know this is not another OPERA?” 5 sigma discoveries at two indepedent detectors will be indisputable.

      • JollyJoker says:

        Phil, what’s your best guess on how much 8 TeV data they’d need to add to the 2011 data to get 5 sigma for both experiments? It should be somewhere around 5.5-6/fb, right?

      • JollyJoker says:

        By both I mean each separately, not both combined

      • Philip Gibbs says:

        I estimate they need to add about 10/fb at 8 TeV, but depending on the roll of the dice it could be anywhere between 5/fb and 15/fb . This is for the combination over all channels for each experiment.

        The significance of the diphoton channel will probably be no better with an additional 5/fb because they were both very lucky with the stats ( or the boson has a much higher branching ratio than expected )

        For the full combination of ATLAS+CMS an additional 4/fb at 8 TeV is probably enough, but again luck comes into play and that is assuming there is really a SM Higgs at 125 GeV.

  9. Tony Smith says:

    Lubos said:
    “… about the Higgs’ being de facto discovered …[at]… 125 GeV …
    Those who disagree don’t really have an opinion …”.

    I disagree and have an opinion – see viXra 1112.0035

    Whether or not Lubos thinks that I should be allowed to have an opinion,
    my opinion is there and the way to evaluate it should be to take more data
    and analyse it properly.

    If my opinion is correct, the July 2012 diphoton data will still not look much like a single peak at 125 GeV.

    The CMS 2011 diphoton data looks like two irregular things around 126 and 132
    while the ATLAS 2011 peaks and valleys do not correspond to the CMS peaks and valleys
    – see the image at
    the ATLAS data from Higgs to ZZ to 4l will continue to show a peak around 240 GeV (and that peak may also show up at CMS).

    Therefore, I look forward to July when data will either support my opinion and its specific predictions (or not).


  10. ondra says:

    CMS crossed 3/fb recorded today! Unfortunatelly last fill was interrupted by cryo failure, lets hope for one more long fill this weekend. It looks like they are figuring the problems with BLMs and injection and they now reached regularly 6.5/nb/s peak lumi with 1.55e11 protons per bunch.

    • Philip Gibbs says:

      They had a good two weeks, and there are two weeks left before ICHEP cut-off. Whatever happens they are going to have good data for the conference.

      • carla says:

        Disappointing to have a cryo outage at this stage, but at least its seems to be down to a faulty PLC. I thought they may have doubled up the PLCs from last year to prevent this sort of problem which was rampant last year and typically led to 2 days of no beams.

  11. wl59 says:

    What they will do now with Mr. Higgs, after it’s better established that his Boson don’t exist ??

  12. Tony Smith says:

    Carla said “… cryo outage … seems to be down to a faulty PLC …”
    with respect to the 27-05-2012 Lost Cryo at P8.

    Now, after Cryo recovery,
    on 28-05-2012 02:28:58 the Elwell comments say:
    “… Beams gone Triplet L8 tripped …”.

    Is this a serious confluence of problems around Point 8 ?
    Could it be related to LHCb (which is at Point 8) ?

    I hope that, whatever it is, it can be corrected soon.


  13. Giorgio says:

    The ATLAS delivered luminosity number 132242.8 in fill 2670 on page is wrong (it is much much too high compared to the CMS number, and it is copied from the previous fill). Do you know somebody at CERN to let him know so that they can correct the error?

  14. Tony Smith says:

    Philip Gibbs said: “… they already cut out ALICE earlier, now they are cutting out some bunches that were only colliding in LHCb but this seems to be for technical stability reasons …”.

    Elwell “… Comments 30-05-2012 18:11:10:
    Beams dumped by ALICE BCM …”.

    Does that mean that they put ALICE back and it caused the dump less than 3 hours after the 15:28:44 fill for physics?


  15. Tony Smith says:

    The LHC (as followed in Elwell comments) seems to put me on an emotional roller-coaster:

    This gets me very high:
    “… Comments 03-06-2012 01:56:24:
    New record fill …”


    this brings me down to low:
    “… Comments 03-06-2012 02:30:23:
    Still no beam for SPS …”.

    On the one hand
    the LHC has delivered around 4/fb for 2012 which should be enough to see whether the 2011 funny peaks around 125 GeV in the digamma channel are replicated (supporting the widely-held (Kane et al) view of 125 GeV Higgs)
    whether the June 2012 digamma data mostly differ from 2011, leaving any realistic consensus to emerge (or not) from data for the rest of 2012.

    On the other hand, if it turns out that another 10 to 15 /fb may be needed to get clarity by the end of 2012,
    things like SPS beam failure are worrisome.


    • ondra says:

      Tony, the very long and very old LHC injector chain is of course worrisome, but for now it didnt cause any major delays.
      I think for future accelerators we should not do something like that but it seems accelerator people like it :). I would actually like to know, why one cant use some short injector which creates given bunch structure and gives some resonable injection energy.

    • carla says:

      You seem to have this idea that accelerators run continuously without a hitch – they don’t. The LHC running around 25% of the time is expected because that’s a figure based upon accelerators like the Tevatron etc. We still haven’t had a lightning strike this year to hit the network and take out the cryogenics for two days yet 😉

      • ondra says:

        Well my point is that while i think you certainly need to gain operational experience with extreme complex machine, some of the errors and mistakes clearly shouldnt be there.

  16. Tony Smith says:

    Carla said (perhaps to me) “… You seem to have this idea that accelerators run continuously without a hitch – they don’t. …”.

    In fact I do not hold any such idea.

    Carla also said “… The LHC running around 25% of the time is expected because that’s a figure based upon accelerators like the Tevatron etc. …”.

    That is not true.
    Here is a quote from Philip Gibbs in his article
    LHC Prospects for 2012 (Prespacetime Journal Sep 2011):

    “… In comparison to other colliders such … the Tevatron a figure of 0.2 is not unreasonable,
    but those machines were always limited to that amount by their design.
    The Tevatron takes a day to build up a store of anti-protons with enough intensity for a single run. In the meantime each store keeps running but the luminosity decreases with a half life of just a few hours. This means that no matter how smoothly it runs, the Tevatron can never do much better than a Hübner Factor of around 0.2.
    The LHC avoids these limitation because it only uses protons which can be extracted from a bottle of hydrogen on demand.
    They also have much better luminosity lifetimes ranging from 10 hours at the start of a run to 30 hours as the luminosity decreases.
    In theory if the LHC can be made to run smoothly
    without unwanted dumps
    and without delays while getting ready for the next fill,
    then they would reach a Hübner Factor of over 0.5. …”.


    • Philip Gibbs says:

      I think that people who are experienced with building accelerators are not surprised by the present efficiency at this stage which is still early days. They may be pleasently surprised by the high luminosities reached already. Many problems have been resolved and they have learnt alot about others. Some further problems may be eliminated by upgrades during the long shutdowns and I imagine that eventually it will run at a higher efficiency than LEP or Tevatron which were limited by the need to build up a store of anti-particles.

  17. ondra says:

    Check the talk of Steve Myers at Physics at LHC conference
    We see that LHC is slightly behind integrated lumi schedule for 2012.
    Btw there are two new conference notes from ATLAS about stop searches with 5/fb,, , but nothing is found. Does that mean that stop rumour has evaporated?

    • Philip Gibbs says:

      I think if something was going to materialise from the original stop rumours we would have it by now, but there is still plenty of room for new stop rumours.

      • ondra says:

        Btw it doesnt look like we will see any 8 TeV results before ICHEP.

      • Philip Gibbs says:

        Right, all the conferences are rehashing 2011 results. However ICHEP will have lots of 8 TeV results including all the Higgs channels and some new ones (H -> Z+gamma) plus some SUSY at 8 TeV

  18. 1. Can the LHC analysis of the data allow us to decipher any nonlinear behavior of the fundamental particles. In other words are the “specific particles” defined by Nature or are the defined by or equations and instruments designed and derived from our equations?

    2. Will renormalization prevent us from ever seeing (observations and/or formulations) of the true nonlinear behavior of Nature?

    3. Is it valid to say that the lowest possible energy of the smallest space (quanta of space) is the potential well that is filled by entropy? Hence, the zero point energy must be greater than potential well threshold in order to have motion (increasing time, evolution). Question is, do particles exist in our universe because the initial conditions at the Big Bang had some lower rate entropy inherent in the process?

    4. A bigger question would be can our visible universe (massive particles) be floating “on top” of a grander, broader universe which energy cannot coalesce into massive particles (dark energy) ?

  19. carla says:

    If the cut off point for ICHEP is 10th June as suggested, then they’ve managed to deliver 5/fb as requested. Great result considering the problems they had at the beginning of this run period. Fingers crossed they should get close to delivering 6/fb by the end of next week so matching last years data set in time for the other summer conferences!

%d bloggers like this: