Anomalous production of Y and W/Z bosons in CDF November 22, 2006Posted by dorigo in news, personal, physics, science.
Between 2001 and 2002 I spent most of my research time in the analysis of associated production of a Upsilon meson and a vector boson in CDF Run I data, together with Giovanni Busetto (then my boss) and Alberto Ribon (a friend who is now at CERN, who was then working in the Herwig group and helped us with Monte Carlo simulations of the processes). The result of the search was a paper we published in 2003:
Search for Associated Production of Upsilon and Vector Boson in p anti-p Collisions at s**(1/2) = 1.8 TeV, Phys. Rev. Lett. 90, 221803 (2003),
The analysis was based on the idea that the associated production process, which in the Standard Model occurs via diagrams such as the one shown on the left, at the not fantastic cross section of fractions of a picobarn (0.5pb for WY and 0.2pb for ZY at the Tevatron), could be enhanced if the pairs were caused by the decay of a Higgs boson (neutral, in the case of ZY, or charged to yield WY), with a mass close to the sum of the boson and meson decay products.
In the Standard Model, Higgs decay does yield a ZY final state (if Mh>MZ+MY), but with a ridiculously low probability – less than 10^-5. So finding the Higgs this way is not thinkable, although the final state of W/Z+Y may be really rich and background-free.
The Y meson can be seen with clarity when it decays to two muons – which it does at the non-spectacular rate of 2.5%. W and Z bosons also yield golden leptonic signatures (electrons or muons), with branching ratios of 22% and 9% respectively. But actually, once you tag the Y in its dimuon decay, you can search for all signatures of W or Z boson production, because the event is already clean from backgrounds. Moreover, the fact that you see the Y signal with clarity in the dimuon mass distribution allows you to estimate – and subtract – all non-Y backgrounds by a procedure we call “sidebands subtraction”: if a background is flat in the mass distribution, by selecting a region close to the Y mass, as wide as your Y mass window cuts, you get a representative set of background events which mimic your signal region but contain no Y by definition (see plot below, from the Run I analysis: the red part of the histogram is the signal region, encompassing the three Y resonances called 1S, 2S, 3S, while the green is the sidebands region).
The analysis we carried out in Padova in 2002 found no signal for the Standard Model process – which was unobservable with the analyzed data of 90 inverse picobarns – and the more so no evidence of any enhanced non-SM production process. We set limits of sigma(WY)xB(Y->mumu) <2.3 pb and sigma(ZY)xB(Y->mumu) <2.5 pb. I must say ours was a very fine analysis, although the results were not interesting at all, our limits being way above SM expectations. However, the novelty of the final state considered won us publication in the prestigious Physical Review Letters.
It is with pleasure that I discovered the other day that two fellow collaborators of CDF are redoing our study with the more than tenfold increased statistics of Run II. Ray Culbertson and Shin-Shan Yu have so far considered only the jet final state of vector bosons, and have not run a Monte Carlo yet for the signal, but they basically redid what we did in 2002. Results are not public yet, but I can tell I was pleased to see the much increased spectra of Y mesons before analysis cuts: the Y family is always nice to behold.
I am eager to know the official results of the analysis, when it will be completed. And who knows, maybe we could see one or two candidates this time, our dataset does allow that now.