A joint Fermilab/SLAC publication

Can the Tevatron find the Higgs? (APS April 2008)


In the past year, there has been a lot of discussion about whether the Tevatron collider at Fermilab can find the Higgs boson before the Large Hadron Collider can. There have been all kinds of claims, and even stories of hints of sightings. In a session today, Brian Winer from Ohio State University gave a very clear presentation about just how close the Tevatron is to potentially finding the Higgs.

Cutting to the chase: Can the Tevatron find the Higgs? Yes, if Nature is kind.

But first let me set the scene.

What do we know about the Higgs now?

The Standard Model of particle physics predicts that a Higgs boson is most likely to have a mass of 87 GeV. (Physicists like to express masses in terms of energies. If you want to convert to an actual mass number, divide the energy by the speed of light squared.) The nature of this prediction is that the Higgs won't necessarily have that mass but it is likely to have it somewhere in a range centered on that number. There is roughly a 2/3 chance that the Higgs will have mass between 60 GeV and 123 GeV.

Experiments at the Large Electron Position collider at CERN showed that the Higgs does not exist at any mass lower than 114 GeV however, and the theory predicts that the mass will be less than 160 GeV with a 95% probability. So physicists are really trying to look in this region of 114 to 160 GeV.

How sensitive do experiments need to be to find the Higgs?

Experiments have different sensitivities to the Higgs depending on its mass. Physicists can estimate how close they are to observing the Higgs by comparing their measurements to theoretical predictions.

At the moment, the performance of experiments is expressed as a factor of how much better the measurements need to be to start to have a chance of being sure that any Higgs-like signal is really a Higgs boson. It is a game of statistics and it gets pretty complicated so I'll just take the simplest possible path through this, warning that the full story is much more detailed.

The biggest problem to start with is that what a collider detector measures is a set of particles, none of which is specifically the particle you are looking for. The Higgs decays into certain sets of particles but other decays look very similar. So when your detector sees a set of particles, you need to make extra sure that you are looking at what you think you are looking at! You can predict how many of certain types of known particles you should see and then look for any excess if you are hunting a Higgs or some other unknown particle.

This "background" of decays from known particles must all be excluded from analysis before you can see the Higgs. The big problem is that the backgrounds are a factor of 100 billion times larger than the signal from the Higgs! Higgs would make up only tens of events out of the trillions measured.

Fortunately, detectors are very good at finding the presence of b and anti-b quarks (they create characteristic jets of particle from the collision). Higgs will tend to decay into b and anti-b quarks along with other particles so the bs can be used to identify most of the collisions you are interested in. In fact, this "b-tagging" can reduce the background by a factor of one billion.

But as Winer put it today, "the first factor of a billion is easy, the last factor of 100 is hard."

How low can you go?

In trying to get this factor of 100 down to a factor of 1, a lot depends on the energy the Higgs has. If the Higgs has an energy at the bottom end of the range, about 115 GeV, and then you combine all the data from both the CDF and DZero experiments, you end up about a factor of 5 short of where you need to be to have a chance at the Higgs. But given that the Tevatron is hoping to collect 4-8 times more data than was used in this analysis, it potentially comes in range though it would still be a real stretch.

If the Higgs has a mass of 160 GeV, then the combined data already brings us to a factor of a mere 1.1 times where we need to be. In other words, a 10% improvement and "Game on!"

Of course, it is not as simple as collecting 10% more data. That would be too easy. That level is really the starting point for when your detectors and dataset have the statistical power to resolve Higgs hiding in the flood of collisions.

But if the Higgs really sits around that mass, then the Tevatron has a genuine chance of finding it soon. Winer said today, "As early as summer, we can start excluding masses around 160 GeV." In other words, the data will be sensitive enough to say either that there is definitely no Higgs at that mass, or it will see signals that look like the Higgs and it will just be a matter of some more data to determine if they really are Higgs bosons.

Is this the Higgs?

And just to tease you a little more, this event could even be a Higgs.

Best potential Higgs event

It is event 6577 from run 196170 and has all the characteristics of a Higgs. It was found on April 5, 2005, at 8:09 a.m. Unfortunately, there is just no way to tell if it is truly a Higgs because the statistics don't allow anybody to prove what this is. But there is some change that you are looking at the first identified Higgs boson!

The possibility of finding the Higgs at the Tevatron is real and the race is really on.

See all posts from the American Physical Society April 2008 conference here.

Latest news articles

Scientists make rare achievement in study of antimatter

Through hard work, ingenuity and a little cooperation from nature, scientists on the BASE experiment vastly improved their measurement of a property of protons and antiprotons.


Scientists observe first verified neutron-star collision

For the first time, experiments have seen both light and gravitational waves released by a single celestial crash.

New York Times

For the first time, astronomers have seen and heard a pair of neutron stars collide in a crucible of cosmic alchemy.


Xenon takes a turn in the LHC

For the first time, the Large Hadron Collider is accelerating xenon nuclei for experiments.