[Paper Review] Study on the azimuthal angle correlation between two jets in the top quark pair production

Kaoru Hagiwara, Junya Nakamura|arXiv (Cornell University)|Jan 5, 2015

Particle physics theoretical and experimental studies48 references3 citations

TL;DR

This study investigates the azimuthal angle correlation between the two hardest jets in top quark pair production at the 14 TeV LHC using merged tree-level matrix elements (up to three partons) with parton showers in PYTHIA8. It demonstrates that including $t\bar{t}+3$ parton matrix elements significantly reduces correlation loss due to jet $p_T$ ordering and rapidity cuts, improving prediction accuracy for CP-sensitive observables.

ABSTRACT

An azimuthal angle correlation between the two hardest jets is studied in the ttbar production process at the 14 TeV LHC. The event samples are generated by merging the tree level matrix elements for the ttbar plus up to 2 or 3 partons with parton showers. The generated event samples show a strong correlation in the azimuthal angle difference between the two hardest jets, as predicted in the analysis based on the tree level matrix elements for the ttbar+2 partons. The effects of merging the matrix elements for the ttbar+3 partons on the correlation are studied in detail. It is found that they play important roles in improving the prediction of the correlation.

Motivation & Objective

• To assess the robustness of azimuthal angle correlations between the two hardest jets in $t\bar{t}$ production under QCD higher-order corrections.
• To evaluate the impact of merging $t\bar{t}+3$ parton matrix elements on the accuracy of azimuthal correlation predictions compared to $t\bar{t}+2$ partons.
• To develop and test a new tree-level merging algorithm that suppresses contamination from lower-parton states in multi-jet samples.
• To quantify the effect of top quark decays on the azimuthal correlation observable, particularly when selecting jets excluding b-jets.
• To support experimental efforts in measuring the CP nature of the Higgs boson by validating simulation tools using $t\bar{t}$-associated jet correlations.

Proposed method

• Event samples for $t\bar{t}$ production at 14 TeV LHC are generated by merging tree-level matrix elements for $t\bar{t}+0$, $t\bar{t}+1$, $t\bar{t}+2$, and $t\bar{t}+3$ partons with parton showers in PYTHIA8.
• A new tree-level merging algorithm is implemented, differing from CKKW-L by improving phase space separation to suppress contamination from $t\bar{t}+0,1$ partons in multi-jet samples.
• The azimuthal angle difference $\Delta\phi = \phi_1 - \phi_2$ between the two highest transverse momentum jets is computed and analyzed as the key observable.
• The effect of top quark decays is modeled by including $t\bar{t}$ decay information directly into the merged matrix element events.
• Jet reconstruction uses the anti-$k_T$ algorithm with varying rapidity cuts to study their impact on correlation loss.
• Comparisons are made between event samples merged with up to 2 or 3 partons to isolate the contribution of higher-parton states.

Experimental results

Research questions

• RQ1 How does the inclusion of $t\bar{t}+3$ parton matrix elements affect the prediction of the azimuthal angle correlation between the two hardest jets?
• RQ2 To what extent does contamination from $t\bar{t}+0,1$ parton matrix elements distort the $\Delta\phi$ distribution in merged event samples?
• RQ3 How do jet $p_T$ ordering and rapidity cuts influence the loss of correlation in $t\bar{t}+2$ parton samples?
• RQ4 What is the impact of top quark decays on the $\Delta\phi$ correlation when the two hardest jets are selected from non-b-jets?
• RQ5 Can the new merging algorithm suppress contamination more effectively than CKKW-L, leading to improved correlation predictions?

Key findings

• The $\Delta\phi$ distribution exhibits a strong correlation between the two hardest jets, confirming predictions from $t\bar{t}+2$ parton tree-level matrix elements even after including higher-order QCD corrections.
• Including $t\bar{t}+3$ parton matrix elements reduces the loss of correlation by up to 5–10% compared to $t\bar{t}+2$ samples, primarily by preventing hard parton shower jets from out-ranking the two hard partons.
• When rapidity cuts are applied, the correlation loss increases significantly in $t\bar{t}+2$ samples due to the removal of one of the two hard jets, but this is mitigated in $t\bar{t}+3$ samples due to the presence of a third hard jet.
• The new merging algorithm suppresses contamination from $t\bar{t}+0,1$ partons more effectively than CKKW-L, though the improvement is not drastic in the $\Delta\phi$ observable.
• The effect of $t\bar{t}$ dilepton decay on $\Delta\phi$ is small when the two hardest jets are selected from all jets excluding the two hardest b-jets.
• The $t\bar{t}+3$ parton matrix elements play a crucial role in accurately predicting the azimuthal correlation, especially under stringent rapidity cuts, and are essential for reducing theoretical uncertainties in CP-sensitive Higgs boson measurements.