2012 CMS Luminosity: 10/fb And Counting!
    By Tommaso Dorigo | August 10th 2012 08:19 AM | 7 comments | Print | E-mail | Track Comments
    About Tommaso

    I am an experimental particle physicist working with the CMS experiment at CERN. In my spare time I play chess, abuse the piano, and aim my dobson...

    View Tommaso's Profile
    The Large Hadron Collider is delivering as expected a large amount of integrated luminosity of proton-proton collisions to CMS and ATLAS, running at a centre-of-mass energy of 8 TeV. The total collected in 2012 by CMS has just now crossed the mark of 10 inverse femtobarns: this is twice as much data as that collected in 2011. And the CMS detector is working like a charm, with all subsystems collecting data flawlessly.

    Physicists need large integrated luminosity to explore rare phenomena, and high energy to probe processes which may only "turn on" above a certain threshold. So given the x2 luminosity so far collected (but we will get to x5 by the end of the year!) and x1.14 energy, the discovery potential of 2012 data is already several times larger than that of 2011 data.

    It is not too hard to figure out why 8 TeV are better than 7: if there existed a new particle with mass of 7.5 TeV, you would never be able to produce it with 7 TeV collisions.... But unfortunately, neither could you with 8 TeV!

    To understand that, first you need to reckon with the fact that while protons collide at 8 TeV (or 7), the hard interaction that gives rise to a potentially new particle creation or new phenomenon never reaches such energy: that is because there are many quarks and gluons (collectively called "partons") in each proton, and what produces the hard collision is not a pair of protons, but a pair of partons; the protons are like soft bags containing hard things. So the chance that the two partons which collide are those carrying the largest fraction of the total energy are slim - indeed the highest-energy collisions we have so far observed involve quarks and gluons whose total center-of-mass energy is in the 3-4 TeV range, and are quite rare.

    So, fine. Say we want to produce a 3-TeV resonance. This is certainly possible with 8-TeV collisions, and less rare than with 7-TeV collisions. But it is still quite, quite rare - only once in several trillion times, or even less depending on what particle you want this to be. So here is why we need more data. If the new particle has a "cross section" of 1 femtobarn, that means we need on average to produce one inverse femtobarn of proton-proton collisions to have a good chance of getting one of them, since the average number of particles produced for a given process given some data is given by the formula

    N = σ L

    where σ is the cross section, and L is the integrated luminosity corresponding to that data.

    Now, one event is usually not enough to prove we have discovered something: backgrounds can always mimic the production of a signal, so we need to collect several events all of the same characteristics to be able to prove that there is a new signal in the data.

    If you will, the above may be taken to mean that a proton-proton machine is more limited by the amount of data it collects than by the total energy of its beams ! This is not entirely correct but it is actually not far from the truth, either.

    To shed more light we can contrast the LHC situation with the operation of the LEP II collider, which accelerated electrons and positrons instead than protons, at up to 209 GeV center of mass energy in the same ring now hosting the LHC, until 2001. Since electrons and positrons are elementary bodies, the collisions deliver all the energy of the projectiles to the hard interaction.

    In a quest to produce the Higgs boson through the reaction e+e- --> ZH, that is associated with a Z boson, CERN increased as much as possible, in steps (inserting new accelerating cavities in the ring where possible) the center of mass energy of LEP II, without caring too much for the amount of data collected at each step: all it mattered was the energy, because the cross section of the searched process was not too small, and every collision delivered exactly the total energy of the machine. We know how the story ended: they could say the Higgs was heavier than 114.4 GeV because they did not see a significant excess of ZH candidates; had the Higgs been lighter, they would have seen an excess. If you sum the minimum mass of an escaped Higgs boson with the mass of the Z boson you get 114.4+91.2=205.6 GeV, close to the actual collision energy -some of the energy does not go into mass but into momentum of the final state products.

    Protons, unlike electrons, are not elementary, but are made of quarks and gluons, so I hope you now understand why the LHC experiments have only just begun their investigation of the high-energy frontier: getting more data will be needed to say we have exhausted the potentialities of the machine.

    There is now a schedule of future LHC operations that looks ahead at least 10 years from now - in the course of this time, CMS and ATLAS will be able to collect hundreds, or even thousands, of inverse femtobarns of data. What new things may appear when we get to analyze that much data, we can only wonder. And of course, the fixing of safety systems in the ring will finally allow the collider to run at 13 or 14 TeV. In the meantime, experiments will require to be upgraded, partly to match the higher number of collisions per bunch crossing of higher luminosity running, which make it harder to reconstruct the trajectory of every particle; and partly to replace systems worn out by high doses of radiation, or to introduce newer technologies in apparata that while still quite advanced and new-looking, but will not stay that for long.


    Amir D. Aczel
    Tommaso, Great explanation of detection power and its relationship with beam energy and luminosity. At a talk I gave yesterday in Boston, many in the audience wanted to know what the chance is, with higher energy and increasing cumulative luminosity, that new physics will be discovered: SUSY or extra dimensions. My own "feeling" is that SUSY is more likely than extra dimensions. (New $3 million prizes from a Russian source--mostly to string theorists--may indicate others may disagree!...). What is your view? Also about LHCb's chance of learning more about matter-antimatter issues.
    Amir D. Aczel
    Dear Amir,

    I have just won a $250 bet (with Gordon Watts, experimentalist in ATLAS) and almost won another of $750 (with Jacques Distler, theoretician in Texas - winning this one will require waiting for analyses to finish, about ten more months) against the LHC finding anything in its first 10/fb of data.

    So I don't believe in either... But I would be among the first to rejoice if we found SUSY! About LHCb, I think (personal opinion) that the chances that we will learn much there are not large either.

    what do you think of the Michael Banks article in Physics World 6 August 2012 saying:

    "... LEP3 … could be constructed within the next 10 years …
    LEP3 would operate at 240 GeV and comprise two separate accelerator rings that would smash electrons and positrons … in the LHC tunnel
    and serve the two LHC's general-purpose detectors - ATLAS and CMS

    a decision would have to be made between building LEP3 and running the high-luminosity upgrade to the LHC in the 2020s

    If it turns out that the LHC finds only the Higgs,
    then Ellis says there would be a strong case for LEP3 …".


    I think it is possible, but not my favourite plan. An extra linear collider could be made instead (with electrons, or a muon collider) to produce Higgs bosons and study them in detail. It depends on the funding we'll get.

    ... Plus, I would say that psychologically it would be a rather tough decision to take.
    Think about it: it would equate to saying "building the LHC was a waste of time - we could
    have stopped LEP II, increased further the beam energy (although with a significant redesign), and get to the same place fifteen years earlier"...

    Nice post. I was wondering, why did they not rather use electron/positrons in the LHC as they seem cleaner and as you say deliver all the energy for each collision?

    Ah, good question. Electrons radiate bremsstrahlung photons at 10^4 times the rate at which protons do, when you try to steer them into a circular orbit. So to keep an electron in the LHC orbit at 46 GeV (==LEP) or 104 GeV (== max of LEP II) you spent a huge amount of power. Going up with the beam energy using electrons and positrons was prohibitive. But as Tony commented above, going up to 120 GeV (which would allow to produce ZH events) is still possible.

    Another attempt would be to build a linear collider. In a linear collider you don't lose energy by bremsstrahlung (if the two beams are just one against the other in a linear configuration) or do so only once during the final turn (as in SLAC).