Wednesday, February 08, 2012

Higgs signal grew from 3.8 to 4.3 sigma

Well, the newer and higher figure is real: the lower one may be from a silly yet pedagogical calculation of a SuperBowl organizers' anomaly by Sean Carroll but it's not far from the obsolete value (around 4 sigma), anyway

From the first half of December 2011, TRF readers have known that a Higgs at 124-126 GeV is a sure thing, despite some nearly unbelievable skepticism from Prof Matt Strassler and 97% of the particle physicists he talks to. ;-) His skeptical blog entry, religiously referring to an impressive yet enigmatic and anonymous majority :-), was later erased and "superseded" by another article which is still nearly unbelievable to me. ;-) Sorry, Matt, I am just not getting it.

At any rate you – and 97% of the particle physicists that your humble correspondent talks to (and Gordon Kane, David Gross, or Nima Arkani-Hamed are just three "far from anonymous" examples you may remember from recent TRF articles) – won't be surprised that the Higgs (or Higgs-like) signal is facing one kind of a future trend only: it's destined to go up.

Phil Gibbs' Higgs combination chart from December 2011. Click to zoom in.

However, you may still be surprised that the signal has been getting stronger even before new collisions are recorded at the LHC (new collisions at 4 + 4 = 8 TeV will begin at the beginning of April). But that's what has occurred, too. Nature blogs described it by the title
Higgs signal gains strength
The informal combination's confidence level has been raised from 3.8-4.1 (or so...) standard deviations to 4.3 standard deviations, largely because of improvements in the CMS analysis.

I think it's not too important to redo all the combination charts because they were convincing almost two months ago and they're even more convincing now. See some recent LHC papers on the Higgs where the older analyses have been refined.

CMS (see new combination paper) continues to prefer a Higgs around 124 GeV, ATLAS (see their new combination paper) sits around 126 GeV. The 2 GeV gap may arguably be explained by a combination of a statistical and systematic error and we'll see later which of the components dominates. However, I am almost totally confident that the 2 GeV gap isn't strong enough to claim a genuine discrepancy between the CMS and ATLAS results, not even at a 2-sigma level.

Still, when you have figures 124 GeV and 126 GeV that are likely to boil down to the same physics, it's damn reasonable to say that the mean value of the mass is somewhere in between, around 125 GeV, even if Tommaso Dorigo tried to prevent you from computing the average. ;-)

The diphoton channel seems to be stronger than expected from the Standard Model in the case of both detectors. The significance of this "excess" could be slowly approaching 2 sigma at this point. Clearly, more data will be needed to become certain on the question whether or not the bumps have "exactly" the size predicted by the Standard Model. However, I think that the reality of the bumps is pretty indisputable at this point.

If you want some numbers, the local significance of the CMS or ATLAS bumps at 124 or 126 GeV is 3.1 or 3.5 sigma, respectively. The root mean square value obtained from 3.1 and 3.5 sigma – ignoring the mass difference – is 4.68. I suppose that 4.3 could be the significance near 125 GeV which is reduced by the mass difference between the detectors, or the reduction from 4.68 to 4.3 has a different subtle origin.

The Higgs signal isn't at 5 sigma yet but it is really not far from it. It's pretty obvious that less than one month of extra data will be enough to strengthen the combined LHC significance level above 5 sigma and claim discovery in a semi-official way. In this context, I would like to ask John Ramsden to prepare those $500 that he will owe me in a few months. Thank you! ;-)

Needless to say, a hypothetical discovery of supersymmetry would be approximately 20 times more important that the discovery of the Higgs. Well, as Sheldon would tell Leonard, it's exactly twenty. The product is equal to $10,000, a number well-known to Jester, a lovelorn ex-boyfriend of SUSY's who later made a big and risky 100-to-1 bet against her. ;-)

No comments:

Post a Comment