Smelling the Higgs

Updates of the Higgs boson searches at the Tevatron will be ready for most of the search channels in time for the winter 2007 conferences, which means in a few weeks… While I have insider information for CDF, I know nothing of the searches in D0.

I of course cannot divulgate what I saw from one presentation of the H->WW search in CDF last week, but one thing I can say – things are getting interesting at the Tevatron!

That particular result, obtained by analyzing events with two charged leptons (electrons or muons) through a test of different hypotheses for the Higgs boson mass with a neural network, will be blessed on February 4th if all is well. But other fresh new results are in the pipeline… So let us take a look at the situation as of last Summer and see what we can expect in the next few months.

This is a spectacularly busy plot (I could have said disgustingly busy, but that’s a detail). However let me try to make sense of the tons of information it contains. The thick red line shows the current combined limit of CDF and D0 searches for the Higgs boson. It shows, for each value of the x axis (a particular Higgs boson mass in the range 100-200 GeV), a limit in the rate of production of that particle, expressed as a multiple of the expected Standard Model production rate.

To be clear, at 160 GeV the rate limit is at 3x the SM prediction, which means that the experiments have excluded with the available data and through searches in all analyzed decay modes an anomalous production of that particle  with a rate over three times larger than expected. At other Higgs masses the limit is less stringent, ranging all the way up to a x10 factor in some less fortunate regions.

But the red line is just the combined result. Let us read the fine print now. Each of the colored lines sketched above the red line is an independent result, obtained by either CDF (full lines) or D0 (dashed ones) in a particular final state. And associated to each line is a number which states how much data was used for the result.

The latter is critical information when trying to extrapolate, without insider information, what the Tevatron could offer at the coming winter conferences. CDF and D0 have on tape almost 2 inverse femtobarns of data now, but are likely to show results based on only a little more than that – looking at the new data is a tough job! So if one were to (optimistically) believe that the experiments will present results based on 1.2 fb-1 in all channels, what could happen to the red line ?

First of all: at very low mass, the results drawn in the plot above are based on very little data: D0 in particular only has up to 0.4 fb-1 there. CDF was a bit faster last summer, but still the WW final state only had 0.36 fb-1, and the ttH only 0.32 fb-1. None of these channels will make a big difference by itself, so if I put them together here is what I find at 115 GeV:

• CDF ZH->nnbb (blue line) will not move down much (was based on 1 fb-1 already)
• CDF WH->lnbb (red line) won’t also change much
• CDF ZH->llbb (dark green line) same thing
• CDF ttH->ttbb (purple line) should move down by a factor of 4 – being a statistics dominated search, the limit decreases almost linearly with increasing data;
• CDF WH->WWW (green line) also should move down by a factor 6 for the same reason;
• D0 WH->lnbb (dashed red line) should go down by probably not more than a factor 2, considering the search is background-ridden and so the limit scales with the square root of the available data (at zeroth order);
• D0 ZH->nnbb (dashed blue line) should go down by a factor 2.5;
• D0 ZH->llbb (dashed dark green line) should go down a factor 4;
• D0 WH->WWW (dashed green line) should go down by a factor 4.
• Then maybe, if D0 did a search in the ttH mode, they could get a similar line as that found by CDF.

Ok, so what would we end up getting at 115 GeV  with 1.3 fb-1 and all available channels squeezed ? I am too lazy to fold together 10 gaussians and compute integrals from it, especially since I would still get a wildly extrapolated guess. I think I am good at eyeballing, so I would guess that the Tevatron would exclude a 3x SM Higgs rate at 115 GeV with 1.3 fb-1 of data analyized in all the mentioned final states.

And what about 160 GeV, the other interesting point in the graph (where the H->WW decay would provide a clear signature, given the reality of the two bosons) ? If I do the same exercise, I get a better result there: probably a 2x SM limit there.

The above guesstimates turn out to be roughly in line with what was predicted the Tevatron would get, back in 2003, by the Higgs Sensitivity Working Group (of which I was a proud member). The low mass limit was predicted to come out slightly better, the high mass one slightly worse. These are features of the data we collected, due to fluctuations of the backgrounds just as much as to imperfect predictions of the HSWG…

Therefore the bottomline is: do not expect that the Tevatron will show significant results at the Winter conferences. Do not get mistaken, pry do go there if you have a chance: the place and the food are usually great, plus skiing in the western alps is wonderful; but don’t go there with expectations to be told where the Higgs is sitting… 

On the other hand, the chance that Nature (the bitch, not the magazine) had something in store at a few times the Higgs cross section is still there! So, keep your eyes open!

Update: I realize that the plot above is too small to show any of the details I am discussing here. So below I attach a zoom-in on the low-mass region.