Phenomenology of NNLO jet production at the LHC and its impact on parton distributions

TL;DR

The paper assesses how NNLO QCD (plus electroweak corrections) jet predictions, including both single-inclusive jets and dijets, constrain proton PDFs in a global fit. By incorporating ATLAS and CMS Run I data at 7 and 8 TeV and exploring scale choices and correlation models, it demonstrates that NNLO corrections are crucial for compatibility with the rest of the dataset and for reducing gluon uncertainties. It also shows that dijet data provide a complementary, perturbatively robust constraint with a stronger gluon impact, though single-inclusive jets contribute more to reducing uncertainty. The study establishes the viability of dijet observables for precision PDF determinations and highlights the need for high-quality correlation handling and future 13 TeV data to further sharpen the proton’s gluon distribution.

Abstract

We present a systematic investigation of jet production at hadron colliders from a phenomenological point of view, with the dual aim of providing a validation of theoretical calculations and guidance to future determinations of parton distributions (PDFs). We account for all available inclusive jet and dijet production measurements from ATLAS and CMS at 7 and 8 TeV by including them in a global PDF determination, and comparing to theoretical predictions at NNLO QCD supplemented by electroweak (EW) corrections. We assess the compatibility of the PDFs, specifically the gluon, obtained before and after inclusion of the jet data. We compare the single-inclusive jet and dijet observables in terms of perturbative behaviour upon inclusion of QCD and EW corrections, impact on the PDFs, and global fit quality. In the single-inclusive case, we also investigate the role played by different scale choices and the stability of the results upon changes in modelling of the correlated experimental systematics.

Figures (19)

• Figure 3.1: The NNLO QCD $K$-factors, Eq. (\ref{['eq:kfactor']}), for the ATLAS 7 TeV (top) and CMS 8 TeV (bottom) single-inclusive jet cross-sections evaluated using NNPDF3.1 PDFs and scale $\mu=\widehat{H}_T$. Results are shown as a function of the jet $p_T$ in different jet rapidity bins, with the central (forward) bins shown in the left (right) plot.
• Figure 3.2: The NNLO QCD $K$-factors, Eq. (\ref{['eq:kfactor']}), corresponding to the ATLAS 7 TeV (top) and CMS 8 TeV (bottom) dijet cross-sections evaluated with NNPDF3.1 PDF and scale $\mu=m_{jj}$. Results are shown as function of the jet $p_T$ in different jet rapidity bins, with the most central (forward) bins in the left (right) plot.
• Figure 3.3: The EW $K$-factors, Eq. \ref{['eq:kfactorEWK']}, for the ATLAS and CMS single-inclusive (top) and dijet (bottom) measurements. For single-inclusive jets the $K$-factors are shown as a function of jet $p_T$ in six different rapidity bins. For dijets they are shown as a function of the dijet invariant mass $m_{jj}$ for different $y^*$ bins for ATLAS (left) or $y_{\rm max}$ bins for CMS (right).
• Figure 3.4: The NNLO QCD $K$-factors for the central rapidity bins of the ATLAS 7 TeV single-inclusive jets (left) and CMS 8 TeV dijets (right), with the Monte Carlo numerical uncertainties shown as filled bands around the central result.
• Figure 4.1: Correlation coefficients between the data of Table \ref{['tab:input_datasets']} and the gluon PDF $g(x)$. Each curve corresponds to a different datapoint, with the value of $p_T$ corresponding to the color code on the right of the plot, and only curves for the points in the largest and smallest rapidity bins are shown. The shaded bands denote regions in which the maximum correlation is greater than 90% of the maximum correlation in the whole plot.