Neutral meson cross-section and direct photon HBT measurements in high energy pp, pA, and AA collisions rely on either subtracting or dividing out the background underlying the $M_mathrm{inv}$ or $Q_mathrm{inv}$ distributions. In this paper we investigate the photon pair distributions for simulated proton-proton collisions at $sqrt{s}=13~$TeV, and show that the correlated part of the background in $M_mathrm{inv}$ gets poorly described by the conventional event-mixing method implemented in experimental analyses. We show that it fails to describe the background as it gives a qualitatively different shape especially at invariant masses around the $pi^{0}$ and $eta$ meson mass. We show that this is caused by the correlations originating from parton fragmentation, and discuss a new method that attempts to describe the correlated part of the combinatorial background by only pairing photons when the parton ancestors are close in $eta$ and $varphi$.
تحميل البحث