Although data from more than 140 trillion proton-proton collisions had now been captured by each detector collaboration, the physicists were still wrestling with only a handful of excess events. And the statistics of the few can be prone to wild fluctuations. Small changes can make big differences.
For example, the statistics of tossing a coin seem very straightforward. We know that there’s a 50:50 chance of getting heads or tails. However, if we look at only a few such tosses, we shouldn’t be surprised if we see a sequence featuring an excess of heads or tails. This doesn’t mean that the coin isn’t ‘true’. It simply means that we haven’t observed enough tosses to give us a representative sample. As we gather more data, we would expect that any excess would gradually disappear.
The results presented in Mumbai did not yet mean that the Standard Model Higgs did not exist. There were still excess events at energies between 115–145 GeV, but this was an energy range that had always been acknowledged to be rather problematic for the LHC.
There was only one thing for it. We would have to be more patient and wait for even more data. Higgs had waited 47 years. Another few months wouldn’t make much difference.
____________
The LHC continued to perform better than expectations through the summer of 2011 and on into the autumn, reaching a peak luminosity of 3650 inverse microbarns per second. The proton run ended on 31 October, with each detector collaboration having amassed more than 5 inverse femtobarns of data from 350 trillion proton-proton collisions.
But these were vaguely troubling times. The experiences at Mumbai had undermined confidence. There had been no announcement on the Higgs from CERN since the Mumbai conference, and no announcement appeared to be forthcoming. The long-promised combined data from ATLAS and CMS was finally released, but this told us nothing new and referred only to the two inverse femtobarns of data that had been available in July. The combined data set was now more than five times as large.
A flurry of excitement was instead caused by the announcement, on 23 September 2011, that a group of physicists working at the OPERA experiment,* buried deep beneath the Gran Sasso mountain in the Appennine range in central Italy, were about to report the results of painstaking measurements of the speed of muon neutrinos generated at CERN, 730 kilometres away. The results suggested that the neutrinos were travelling through the earth and reaching their destination very slightly faster than the speed of light.
As the debate about faster-than-light neutrinos evolved, other CERN physicists were busy trying to explain how non-discovery of the Higgs still represented an important step forward for high-energy physics. It was certainly true that non-discovery would undermine the Standard Model and have theorists scrambling back to the drawing-board. But, with the best will in the world, finding nothing is simply not the same as finding something.
With the outlook rather gloomy, an announcement that the CERN Council would hold a meeting with member state representatives to discuss the latest developments in the search for the Higgs appeared largely underwhelming. The first day of the meeting, scheduled for 12 December 2011, would be closed. But public talks from Gianotti and Tonelli scheduled for the next day looked a little more promising. Was there something interesting to tell, after all?
The world’s media gathered at CERN on Tuesday, 13 December. Journalists were no doubt somewhat bemused by the rather dry, technical presentations they witnessed, but the conclusions were nevertheless quite compelling.
Combining the data from several different possible Higgs decay channels, the ATLAS collaboration had observed an excess of events corresponding to 3.6-sigma above the predicted background for a Higgs boson with a mass of 126 GeV. CMS reported a combined excess of events with slightly lower statistical significance of 2.4-sigma for a Higgs with a mass around 124 GeV.
The physicists nevertheless urged caution. ‘This excess may be due to a fluctuation,’ said Gianotti, ‘But it could also be something more interesting. We cannot conclude anything at this stage. We need more study and more data. Given the outstanding performance of the LHC this year, we will not need to wait long for enough data and can look forward to resolving this puzzle in 2012.’16
Heuer explained: ‘[The data provide] intriguing hints in several channels in two experiments, but please be prudent. We have not found it yet. We have not excluded it yet. Stay tuned for next year.’17 Jon Butterworth told Britain’s Channel 4 News: ‘We’re all rather excited because it looks very suggestive and, as Rolf Heuer said, it’s turned up in a few different places at once. But we still need to roll the dice a few more times.’18
Higgs himself echoed this party line: ‘Ah well, I won’t be going home to open a bottle of whisky and drown my sorrows, but equally I am not going home to crack open a bottle of champagne either!’19
In a blog entry posted the same day, Dorigo declared the results to be ‘firm evidence’ for a standard model Higgs boson with a mass around 125 GeV.20 There followed a short but intense war of words in the blogosphere as American theorist Matt Strassler adopted a more conservative view, arguing that Dorigo’s use of the word ‘firm’ was unwarranted: ‘If he had said “some preliminary evidence” he would have gotten away with it. As it is, it seems to me that he has crossed a line…’21
In truth, the physicists were collectively urging prudence, but many were individually ready to take a gamble, as Butterworth explained to me: ‘We really do need data to be sure, but I would bet on this myself. [It] depends how much of a betting man you are.’22
At the very least, there were good grounds for optimism. With the LHC scheduled to re-start proton physics in April 2012, the focus of attention would once again turn to the big summer conferences.
____________
The parameters for the next physics run at the LHC were decided at a workshop in Chamonix in February 2012. After a year of highly successful operation, the engineers were now much more confident about the machine’s capabilities, and agreed to push the total proton-proton collision energy to 8 TeV. This higher energy could be expected to offer up to thirty per cent enhancement of the rate of Higgs production which, when the effects of increased backgrounds were taken into account, still resulted in a 10-15 per cent increase in sensitivity. The target was set to collect 15 inverse femtobarns of data at this higher collision energy during 2012. This would surely be enough data finally to end the search for the Higgs.
On 22 February, it was revealed that the OPERA results implying faster-than-light neutrinos were in error. A loose fibre-optic cable had caused a slight delay to the timing measurements, which had translated into a decrease in reported flight-time of the neutrinos of about 73 billionths of a second. When corrected, the measurements were entirely consistent with neutrinos travelling at light-speed.
To a large extent, this was an embarrassing conclusion to the saga, but physicists everywhere breathed a sigh of relief, secure in the knowledge that Einstein’s special theory of relativity was safe. A couple of high-profile members of the OPERA collaboration resigned their positions. This was a rather sobering reminder (if one was needed) of what can happen when an elaborate physics experiment makes some very public announcements that are subsequently shown to be wrong.
Operations at the LHC were re-started on 12 March and the collision energy of 8 TeV was achieved eighteen days later. The proton physics run began in earnest in mid-April. Instantaneous luminosity built to a peak of 6760 inverse microbarns per second. Although data gathering was slowed slightly by some technical problems associated with the cryogenics, by the end of May the LHC was delivering an impressive 1 inverse femtobarns of data per week to each detector collaboration.
Strong momentum was now building for an announcement at the 36th International Conference on High Energy Physics (ICHEP), scheduled to commence on 4 July in Melbourne, Australia. By 10 June, the notional cut-off date beyond which it would not be possible to analyse further data in time to be presented at the conference, the LHC had delivered about 5 inverse fem
tobarns to both ATLAS and CMS, as much data as had been gathered through the whole of 2011.
Inevitably, rumours began to appear in the high-energy physics blogs. Peter Woit reported a rumour suggesting that strong hints of the Higgs were again being seen, with data from 2011 and about half the data available for 2012 showing an excess of events with 4-sigma significance in the H →γγ channel. Speculation grew in intensity. All the signs were that both ATLAS and CMS might report data showing excesses just short of the 5-sigma needed to declare a discovery. If this were really the case, then there could be little doubt that combining the results from both collaborations would likely tip the conclusion in favour of something like the Higgs.
But would the collaborations take this step? If they didn’t, then the matter would remain officially unresolved until yet more data had been obtained. This would leave the bloggers free to publish their own quite reasonable but definitely unofficial data combinations. The bloggers could find themselves possibly declaring a ‘discovery’ that couldn’t be officially sanctioned. There was no precedent for this situation in the entire history of science.
And then, in a surprise move, CERN announced that it would hold a special seminar at the laboratory in Geneva on 4 July, as a ‘curtain-raiser’ to the ICHEP conference. The seminar would provide updates on the search for the Higgs from ATLAS and CMS, to be followed by a press conference. Higgs, Englert, Guralnik, Hagen and Kibble were all invited to attend.*
Surely, this was a sign that one or both detector collaborations had achieved the 5-sigma significance required to declare a discovery? The speculation intensified. Not to be outdone, Fermilab physicists reminded us that each of the two Tevatron collaborations, D0 and CDF, had accumulated almost ten inverse femtobarns of data at a lower collision energy. At a conference held in Moriond, France, in March physicists from Fermilab had revealed results suggesting a 2.2-sigma excess in the range 115-135 GeV, with emphasis on decays to two bottom quarks, a channel not easily observed at the LHC because of high background. In a subsequent seminar on 2 July, two days before the CERN announcement, Fermilab physicists declared that through improvements in their analysis they had pushed the significance to 2.9-sigma. Of course, this was insufficient to declare a discovery, but would certainly provide strong corroboration for any subsequent discovery announcement.
____________
On 4 July I watched the live webcast from CERN from the comfort of my office, and tracked the audience reaction through blog entries posted live by Dorigo, who was present at the seminar.
Heuer declared this day to be special for several reasons. This was, after all, the opening event of an international physics conference, the first such conference to be opened via video link from a different continent.
First up was Joe Incandela, professor of physics at the University of California, Santa Barbara, acting as spokesperson for CMS. He seemed nervous, as if aware of the historic importance of the stage on which we was now standing, at the centre. His nervousness eased as he got into his stride.
His presentation justifiably made much of the bewildering complexity of these experiments. Simply to summarise the outcome in terms of a single result – the answer to the Shakespeare question – would not respect the efforts of all involved in running the LHC, operating the detectors, setting the triggers, managing the pile-up of events, calculating the background, managing the worldwide computer grid, performing the detailed analysis, and not sleeping much. Incandela spent quite some time on these technical aspects, as though to reassure everyone that there could be no doubt about the results that he was about to reveal.
When he finally got to it, his punchline was thrilling. Combining the 7 TeV collision data from 2011 and the 8 TeV data from 2012 had produced an excess of events near 125 GeV in the H →γγ channel with a 4.1-sigma significance. A similar combination of data for the H → Z0Z0 → +-+- channel had yielded an excess of events with 3.2-sigma significance. Putting the data for these two channels together gave a 5.0-sigma excess. The excess expected of a Standard Model Higgs boson at this mass is 4.7-sigma. ‘It’s nice to be at five,’ Incandela said.23
The room erupted in spontaneous applause.
There were some further results to report, relating to other decay channels, but these added little to the overall picture. The combined results are shown in Figure 27(a), in terms of the ‘p-value’ – a measure of statistical significance of the results – vs. the Higgs mass.
Now rather pressed for time, the seminar swiftly moved on to the second detector collaboration. Fabiola Gianotti stood to present the ATLAS results. She covered much the same ground, emphasising important technical aspects of the experiment. I was struck by one singular fact: with a total of 10.7 inverse femtobarns of data, the number of 126 GeV excess events that could be expected in the H →γγ channel was estimated to be just 170. The number of background events at this same energy was expected to be 6340, a signal-to-background ratio of just three per cent.
Gianotti’s punchline was much the same as her CMS colleague. Combining the 2011 and 2012 data had produced an excess of events at 126.5 GeV in the H → γγ channel of 4.5-sigma, a significance somewhat larger (by a factor of two) than the Standard Model prediction. Corresponding data for the H → Z0Z0 → +-+- channel produced an excess at 125 GeV of 3.4-sigma significance. Combining the data for these two decay channels gave a 5.0-sigma excess, compared to a Standard Model prediction of 4.6-sigma. The results are summarised in Figure 27(b).
FIGURE 27 Preliminary results reported by the CMS and ATLAS collaborations on 4 July 2012. These plots show the variation of ‘p-value’ – a measure of statistical significance – vs. Higgs particle mass. (a). The CMS results show excess events for the H →γγ and H → Z0Z0 → +-+- channels and the combination of these, which reaches the all-important 5-sigma level. The dashed line shows the excess predicted for a Standard Model Higgs. (b). A similar plot from ATLAS shows much the same result.
Source: © Copyright CERN
Both collaborations had found the 5-sigma evidence sufficient to declare a discovery. More applause.
Heuer declared: ‘As a layman I would say that I think we have it. Do you agree?’24 There could be little doubting that something very much like the Standard Model Higgs boson had been discovered and, to the layman, this was indeed ‘it’. But the physicists have more exacting standards. They were now rather cagey about what precisely what kind of discovery they had just announced and, under gentle prodding from journalists in a subsequent press conference, stuck to the conclusion that this new particle was consistent with the Higgs. They refused to be drawn on the question of whether or not this indeed was the Higgs.
The simple facts are that the new boson has a mass of between 125-126 GeV and interacts with other Standard Model particles in precisely the way expected of the Higgs boson. Apart from the observed enhancement in the H → γγ decay channel, the new boson’s decay modes to other particles have the ratios expected of a Standard Model Higgs. Whilst the ATLAS and CMS experiments are clear that this is a boson, neither is clear on the precise value of its spin quantum number, which could be 0 or 2. However, the only particle anticipated to have spin-2 is the graviton, the purported carrier of the force of gravity. Spin-0 is therefore much more likely. Paraphrasing Rubbia, we might be tempted to declare, with some justification: ‘It looks like a Standard Model Higgs, it smells like a Standard Model Higgs, it must be a Standard Model Higgs.’
In truth these results represent a critical milestone on another long journey. A new boson has been discovered that looks to all the world like a Higgs boson. But which Higgs boson? The Standard Model needs just one to break the electro-weak symmetry. The Minimal Supersymmetric Standard Model demands five. Other theoretical models make other demands. The only way to find out precisely what kind of particle has been discovered is to explore its properties and behaviour in further experiments.
The CERN press release commented:25
Positive identification of
the new particle’s characteristics will take considerable time and data. But whatever form the Higgs particle takes, our knowledge of the fundamental structure of matter is about to take a major step forward.
The seminar closed with some thoroughly deserved backslapping and self-congratulations. When asked for his observations, Peter Higgs congratulated the laboratory on its remarkable success and said: ‘It’s really an incredible thing that it’s happened in my lifetime.’26
An important chapter in our efforts to understand the basic nature of material substance is drawing to a close. Another, exciting new chapter is about to begin.
EPILOGUE
The Construction of Mass
What is the world made of?
In the mid-1930s we would have explained that all the material substance in the world is made of chemical elements and that each element consists of atoms. Each atom consists in turn of a nucleus composed of varying numbers of positively charged protons and electrically neutral neutrons. Surrounding the nucleus are negatively charged electrons, bound by the force of electrical attraction. Each electron can take either a spin-up or spin-down orientation and each atomic orbital can accommodate two electrons provided their spins are paired. Electrons may move from one orbital to another through the absorption or emission of electromagnetic radiation in the form of photons.
We would have explained that the weight of the 18-gram cube of frozen water in the palm of your hand is derived from the collective mass of 10,800 billion trillion protons and neutrons.
Today our answer has become considerably more refined.
The protons and neutrons in the nucleus are not, in fact, elementary particles. They are composed of fractionally charged quarks. A proton consists of three quarks of different ‘flavours’ – two up and one down. The quarks are also distinguished by their ‘colour’: red, green, and blue. The two up-quarks and the down-quark in a proton all have different colours, the resulting combination appearing ‘white’. A neutron consists of one up-quark and two down-quarks, again with each quark taking a different colour.
Higgs:The invention and discovery of the 'God Particle' Page 17
