UPDATE--There is a very nice summary of the Higgs, what is expected and a liveblog during the talks at http://blog.vixra.org/
Tomorrow morning (at 8:00 Eastern time), an extremely important announcement will take place at CERN, in Geneva. The entire field of particle physics is on pins and needles. This announcement could be the most important in 30 years.
In three earlier diaries, I discussed the status of the Large Hadron Collider's (LHC) search for the "holy grail" of particle physics, the Higgs Boson. The first diary, in July, described what the Higgs boson is and why it is so important. The second diary, a week later, gave the first results presented from the LHC. No Higgs boson was found, but the range of possible masses was narrowed. Four weeks ago, the final update on all of the data collected through August, was presented in the third diary. All three diaries got a lot of comments and made the rec list.
From August to the end of October, four times as much data was collected as previously, and there was clearly enough data to cover the entire region in which the Higgs might hide. I stated that it was likely that the results would be out in late February or early March. But apparently the results are sufficiently exciting that the experimenters will be releasing results now. I know reasonably well what the results will be (it's hard for two experiments with 3000 physicists each to keep secrets -- although it must be emphasized that until the announcement, these are just rumors) and will discuss them below. You will not learn what the Higgs boson is here---go to the first diary link above to learn about it (or look at some of the links given by DarkSyde Saturday morning).
For 30 years I have been waiting for this. And now, in honor of our European colleagues, I invite you to go below the Great Orange Croissant for the details.
The CERN experimenters will not be formally announcing the discovery of the Higgs boson tomorrow. Instead they will probably announce "evidence" for it. But the evidence is strong, and could be just below the threshold for formal "discovery".
When a Higgs is produced at the LHC, it decays immediately into a shower of particles. But every signature of a Higgs decaying also can arise from events that have nothing to do with a Higgs. So you need a lot of data. An example is the following: Suppose you have a theory that a coin will give heads 53% of the time. If you throw it 100 times and get 53 heads, that does not prove your theory is right, since normal coins will often get 53 (or more) heads in 100 tosses. If you get 50 heads, it doesn't prove your theory is wrong (since a normal coin will sometimes give 47 heads). You simply need many, many more tosses to test the theory. By the end of October the LHC had collected enough data to really find out if the Higgs is present.
So what will they announce? All of the properties of the Higgs are known except its mass. Before this year, the Higgs mass was thought to be between 114 GeV and 500 GeV. Three weeks ago, the experimenters announced that they had excluded the possibility of a Higgs between 144 and 500 GeV, basically leaving a small window left between 114 GeV and 144 GeV. As noted in earlier diaries, however, for decades particle physicists (for reasons discussed in those diaries) have expected the Higgs to be between 114 GeV and 135 GeV. Thus, quoting Robert Garisto in the NY Times last month, the "LHC has shown that the Higgs is not where we thought it wasn't". It has also been known that the data collected by the end of October would be enough to explore the remaining window. They are announcing some of the results from that full collection of data.
There are two experiments at the LHC, called ATLAS and CMS. They are very different, with completely different detectors. They search for two possible decays of the Higgs. If a Higgs is heavier than 130 GeV, it mainly decays into two W or Z bosons (well studied and understood particles associated with the weak force). If it is below 130 GeV, it mainly decays into b-quarks, but those are swamped by huge backgrounds. So below 130 GeV, one must look for the decay of the Higgs into two photons (gamma rays). This only occurs once in every thousand Higgs decays, but is very clean--we understand photons better than anything.
Schematically, the Higgs decay looks like this:
So one looks for two photons with a total energy precisely equal to the rest energy of the Higgs (i.e. its mass). As noted earlier, though, many processes produce photons, but over a range of energies:
On the left, you see the background from these other processes. If there is a Higgs, more will be seen around the Higgs mass, giving a bump that you see on the right. How big does a bump have to be to claim that a particle has been discovered? One must be very careful---there will be bumps and dips even if no Higgs is there due to random statistical fluctuations. In the case of the Higgs, there will be 400-500 non-Higgs photons for every Higgs-related photon, so the bump is quite small. Just as with the coin, you need a lot of data to establish a signal.
When scientists report an experimental result, they report both the result and the "experimental error". When looking for a bump over a background, there is always the chance that the background will happen to fluctuate up (like a fair coin giving 53 heads). In describing the significance of a result, physicists use a quantity called "sigma". This is familiar to statisticians-- for any experimental result, there is
A 68% chance of being within one sigma of the actual value
A 95% chance of being within two sigma of the actual value
A 99.7% chance of being within three sigma of the actual value
A 99.9999% chance of being within five sigma of the actual value.
With the Higgs search, one compares to the no-Higgs hypothesis. If a Higgs is there, there will be a discrepancy with the no-Higgs hypothesis. With only a little data, it will be a small discrepancy, and might be only one sigma (which could just be a statistical fluctuation). As more data comes in, sigma gets smaller, and the number of sigma in the discrepancy will grow, providing more and more evidence that a Higgs is there.
You might think that three sigma (99.7% chance) is close enough, but errors are sometimes underestimated, and the physics community long ago decided that a three sigma deviation would be considered "evidence" and a five sigma deviation would be needed for "discovery". I know of a few three sigma announcements that have faded away, but five sigma (note for experts--this assumes there are no large systematic errors) do not generally go away. In addition, one must be very careful about the "look elsewhere effect"---if you were to measure 300 things, then for one to deviate from expectations with a 1/300 probability would not be unusual, and would even be expected.
Tomorrow, the rumors say that ATLAS will announce a huge excess in photons, indicating that a Higgs particle of mass 126 GeV is decaying into two photons. The uncertainty in the number 126 will be about 2-3 GeV. It will be a 3.5 sigma announcement, which has less than a chance in a thousand of being a statistical fluctuation, and constitutes the first evidence for a Higgs. The CMS experiment, according to rumors, will also announce a huge excess in photons, indicating a Higgs particle of mass 124-125 GeV (again with an uncertainty of about 2-3 GeV). It will be a 2.5 sigma announcement. Strictly speaking, the two experiments must do a lot of work to combine the results (they have, to date, been completely independent). But one can get a crude estimate of the combined result by combining the sigmas "in quadrature". Two numbers, a and b, added in quadrature give the square root of a^2 + b^2. For 3.5 and 2.5, that gives 4.3. So the combined result will be about 4.3 sigma, just short of formal discovery. Again, a word of caution---the "look elsewhere effect" discussed above will lower the significance somewhat (my back-of-the-envelope gives about 3.8 sigma).
The combined result will be announced in February or March. The LHC starts up again in March, and a couple of months later the signal will exceed five sigma and formal discovery will, hopefully, be announced. For the first time since Newton came up with the concept of mass, we will understand the origin of the masses of elementary particles.
There is still much to do. With next year's run, they should be able to see decays into W's, Z's and possibly into b-quarks. All are needed to confirm the Standard Model. And many extensions of the Standard Model (including the most popular, called supersymmetry) have several Higgs bosons, and the LHC will begin to be sensitive to those particles as well. The new era has really begun.
If this diary is still active at 8:00 in the morning, I will liveblog the talks here. They are quite technical, but I'll try to give an idea of what's going on.
I've decided not to liveblog the talk. The following comment is below:
Live-blogging seems to be not to be particularly useful on a site like this--one really needs to see the figures and talk about them first. I'm also not good at copying figures down quickly. So I don't think I'm best positioned to live-blog the results. CERN will issue a press release at 6:20 am PST, and that will be more understandable.
You can see the talks online at http://public.web.cern.ch/...
Better, it will be live-blogged at
http://www.quantumdiaries.org/...