Jet and track veto

 Events with Higgs particles produced through vector boson fusion can be separated into four regions centred around the two tag jets and the two vector bosons in the decay as shown in fig. 8.2 and fig. 8.6. To the lowest order, there will be no colour flow between the regions, as they are not connected by any quarks or gluons carrying colour charge. The lack of colour flow between the jets developing causes the number of tracks outside the four regions to be low.

For QCD type events created through quark and gluon fusion processes there will in general be colour flow between all the jets in the events, creating particles in the regions between the jets as well.

Vetoing events with a large number of particles in the central region and outside the two large-cone jet can be used to reject backgrounds like t$\bar{t}$, jW and QCD-jet events which all have colour flow between the final state jets. Higher order QCD processes will to some extent disturb the clean picture described above as the partons produced in the higher order interaction can connect the outgoing jets in the vector boson fusion process and thus create a colour flow between them. The higher order QCD processes are not directly simulated by PYTHIA but are instead effectively included in the initial and final state radiation. This approximation causes a quite large uncertainty in the power of a track counting veto.

Up to 90% of all vector boson fusion events will have colour flow between the different final state jets [78], but in most cases the Q ² transfer in these colour strings will be low and only allow few particles with relatively low p_T to be created. Counting only particles between the jets above a p_T threshold, has thus the potential to increase the power of a track veto.

A track counting veto is not trivial at high luminosity LHC conditions. Only charged tracks can be reconstructed and only in the central pseudorapidity region |$\eta$| < 2.5 . That it is possible was demonstrated in section 7.4, where it was shown with a full detector simulation that tracks with transverse momentum of a few GeV could be assigned to specific primary vertices with a high efficiency at high luminosity.

In fig. 8.8 is shown the distribution in the number of charged tracks, with p_T above 3 and 4 GeV respectively, for the Higgs signal and the t$\bar{t}$ background after all other identification cuts. Only tracks outside the large-cone jet in the central region are counted.

  

Figure: The distribution for the Higgs signal (boxes) and the t$\bar{t}$ background (crosses) in the number of tracks with |$\eta$| < 2.5 and outside the jet in the central region. The cut on transverse momentum was 3.0 GeV in (a) and 4.0 GeV in (b).

The Higgs events with regions empty of particles can also be identified by looking for low energy jets in the calorimeter outside the high energy central jet. The method has the advantage compared to the track counting method that it registers neutral tracks as well. However, there is no way in the calorimeter to distinguish between a jet from the pile-up or a jet from the hard interaction, which makes the method sensitive to pile-up unless a quite high threshold is used on the transverse energy of the required jets. In fig. 8.9 is shown the distribution in the number of jets with |$\eta$| < 2.5 and a transverse energy cut on the jet energy of 20 GeV respectively 50 GeV for the Higgs signal and the t$\bar{t}$ background.

  

Figure: The distribution in the number of jets with |$\eta$| < 2.5 in the Higgs signal (boxes) and t$\bar{t}$ background (crosses). A transverse energy threshold of 20 GeV (a) and 50 GeV (b) was used.

The effect of a jet veto is summarised in table 8.6 and the effect of a track veto in table 8.7. A tracking efficiency of 95% has been taken into account. On average the pile-up will at high luminosity have 3.5 (0.9) charged tracks with p_T > 3(4) GeV. To allocate the few tracks from the pile-up to primary vertices different from the Higgs vertex will not be a problem. For this reason pile-up is not considered to degrade the track veto significantly.

 

$$\bar{t}$$ Jet veto (GeV) H eff. (%) jW eff. (%) S/N

E_T > 20 85.3 2.44 15.6 4.4

E_T > 30 92.0 3.48 18.9 3.8

E_T > 50 96.9 6.97 33.6 2.1

 

 

$$\bar{t}$$ Track veto (GeV) H eff. (%) jW eff. (%) S/N

p_T > 3 77.9 < 1.0 11.7 6.8

p_T > 4 66.8 < 1.0 7.0 8.0