Endgame for the Tevatron

With little fanfare, last week the Tevatron at Fermilab, and the two experiments CDF and D0, emerged from an 11-week shutdown for what will likely be the final run of the collider, which is over 20 years old. In the past year, the machine has regularly set new records for luminosity (essentially the number of collisions per second) and delivered over 2 fb^-1 (inverse femtobarns) of proton-antiproton collisions at a center of mass energy of 1.96 TeV to the experiments, still the highest in the world. The startup has gone very smoothly, and the Tevatron delivered a solid load of data to the experiments last week.

This funny unit, inverse femtobarns, allows us to calculate how many collision events of a certain type to expect. Take top quark pair production, for example. For protons colliding with antiprotons at Tevatron energies, we can calculate (or measure) what we call the production cross section. This cross-section is in fact expressed as an area, a very small area, since protons and antiprotons are so small. One “barn” is 10^-28 m², and the cross section for top pair production is about 7 x 10^-12 barns, or 7 pb. By multiplying the cross-section times the integrated luminosity we can get the number of top quark pairs events produced. To get the number we actually observe in the detector, we need to take into account the efficiency for reconstructing them.

The plot here shows the history of what we call Run 2 at the Tevatron. After a very slow start in 2001, following a five-year shutdown to upgrade the whole complex, the collider set new luminosity records year after year, and has nearly delivered 7 fb^-1. This final run is expected to last two years, until the end of 2011, by which point we hope to have recorded another 5 fb^-1, nearly doubling the present sample.

This past year had been expected to be the year of the LHC at CERN. But the magnet quench incident of one year ago caused a delay of over a year in repairs and retrofits. It is still expected that the LHC will return to commissioning in November of this year, possibly colliding protons on protons before the end of the calendar year, albeit at low energies. When it does come online at higher energies, probably early next year, it is expected that the LHC will deliver no more than about 0.2 fb^-1 at a collision energy of 10 TeV, five times that of the Tevatron. Even with such a small sample, there could be striking discoveries at the LHC which are out of reach for the Tevatron, simply because the LHC energy is so much larger.

The media love this sort of race, and have portrayed it as a race to discover the particle the media has heard the most about – the Higgs boson. With 12 fb^-1, even combining all the search modes and channels, and combining the data from both experiments, a standard model Higgs boson might be seen at the three standard deviation level, but almost certainly not the five standard deviation level, which is the gold standard in the field. The LHC won’t be able to see a standard model Higgs boson with the initial sample either. It will take a year or two at higher luminosity, probably starting in 2012, to get there.

To my mind, if there is a race, it is a race for the unknown. What I worry about, what I literally lie awake thinking about, is whether we are looking at the Tevatron data exactly the right way. People have searched for many different new physics signals at the Tevatron, but there has been no unambiguous observation of anything beyond the standard model. To get a five sigma discovery with the remainder of the Tevatron data, it would have to be the case that there is already about a three sigma excess in the data we have. But have we looked at everything?

Nevertheless, a hard-core of dedicated, talented, and very new-physics-hungry physicists will continue to operate the detectors and analyze the data to come, myself among them. Like many, I am playing both sides: when LHC data come we’ll analyze that, too.