Mohamad-Ziad Charif - Antares

Figure 6.30: Comparison of the Nhit distribution of data, MC: atm µ &atm ν , andtwo dark matter models ( 50 GeV b + b − & 1000 GeV τ + τ − ) which were normalizedto the number of events of atm ν . A cut of Λ >-5.6 was applied here.The muon and neutrino Monte Carlo are normalized to the lifetime of the datataking.XSIn figure 6.31 we can see the distribution of cos(θ) with a cut of Λ >-5.6 .Again the agreement is pretty good between data and our background MC, theloose cut of Λ ensures that atm µ dominates our data sample.Finally in figure 6.32 we can see the distribution of β. This variable is an estimationon the uncertainty on the direction of the reconstructed muon. This variableis a good separator of atm ν from atm µ . However, we can see that the distributionof this variable changes as a function of the dark matter model as well. HeavierWIMPs who have harder spectra seem to peak higher and drops quicker whencompared to lighter WIMPs with softer spectra (1000 GeV τ + τ − Vs. 50 GeVb + b − ). We would expect when β is used in the optimization process to play therole of another discriminator between background and signal (in addition to Nhitand ∆(ψ)) and heavier WIMPs would have a tighter cut on β when compared tolighter ones, and the same with harder when compared with softer spectra.139