Skip to main content

Parton Distribution Function determinations for the future of particle physics phenomenology

Periodic Reporting for period 1 - NEXTGENPDF (Parton Distribution Function determinations for the future of particle physics phenomenology)

Reporting period: 2015-09-01 to 2017-08-31

The Standard Model encodes our present knowledge of high energy particle physics. Developed over 40 years, starting from the 1960s, the Standard Model has proven remarkably successful in describing observations from a wide variety of experiments carried out at different particle accelerators, spanning a wide range of increasing energies. The discovery of the Higgs boson by the ATLAS and CMS experiments at the Large Hadron Collider has completed the search for the particles predicted by the Standard Model.

Parton Distribution Functions, which encode the information on how quarks and gluons are bound inside hadrons, are one of the fundamental ingredients of theoretical predictions for observables at hadron colliders. Indeed they are often the dominant source of uncertainties on predictions for precision observables at the LHC.

In particular, the NEXTGENPDF project is rooted in the development of the so-called NNPDF methodology for the determination of Parton Distribution Functions, which makes use of advanced statistical methods and artificial intelligence techniques.
The pillars of the NNPDF methodology are:
- the use of a large dataset based on results from different experiments, including the most recent PDF sensitive results from the ATLAS, CMS and LHCb experiments at the LHC;
- the use of the most advanced theoretical predictions available for the observables included in the analysis;
- the use of advanced statistical learning methods (for example Artificial Neural Networks and Monte Carlo techniques) for performing an unbiased and reliable determination of Parton Distribution Functions and their associated uncertainties.

The goal of the NEXTGENPDF project is to make use of all the relevant experimental information from the LHC experiments, the most accurate theoretical predictions, innovative techniques for estimation of theoretical uncertainties and advanced statistical inference methodologies to deliver new sets of Parton Distribution Functions that match the precision requirements of the experiments at the LHC Run II and at future colliders, maximizing their potential for discoveries of new physics.
In the context of the NEXTGENPDF project we worked on the refinement of the so called NNPDF methodology for the determination of Parton Distribution Functions (PDFs), a fundamental ingredient of theoretical predictions used in analyses of the data coming from the Large Hadron Collider (LHC) experiments.

In particular, we studied the impact of including new data coming from the LHC experiments in parton distribution function determinations. We also extended the theoretical framework of parton distribution functions determinations by studying the impact of parametrizing the charm content of the proton.

Finally we delivered a new set of parton distribution functions (the NNPDF3.1 set) that is arguably the most up-to-date determination of parton distributions to be used by the LHC experiments and by theorists in precision phenomenology studies and searches for new physics beyond the Standard Model,
The NNPDF Intrinsic Charm fits are the first determination of Parton Distribution Functions that include a consistent treatment of the charm content of the proton. This is achieved by parametrising, for the first time, the charm Parton Distribution on the same grounds of the light quark and antiquark and gluon distributions. Fitting the charm distribution removes one of the main biases which still affect most PDF determinations thus providing a more reliable determination of their uncertainties, in particular for the study of processes that involve directly the charm quark, like the production of a Z vector boson in association with a charm quark at the LHC.

We studied for the first time the inclusion of recent ATLAS and CMS measurements of the Z boson transverse momentum distribution in Parton Distribution Function determinations. These data provide a strong constraints on the quark and gluon PDFs in a region that is, at the same time, poorly constrained by other data and relevant for Higgs boson production at the LHC. For the first time, we made use of recent computations of the Z transverse momentum distribution that, for the first time, computed this observable at high accuracy in the theory of strong interactions. In combination with the use of the NNPDF methodology, this allowed us for the first time to really assess the potential of including these data in a global Parton Distribution Functions determination.
NNPDF3.1 Parton Distribution Functions
Proton proton collision in the ATLAS detector at the LHC (Copyright - ATLAS Collaboration)