Bayesian reweighting of nPDFs

Bayesian reweighting of nuclear parton densities

We have released a work to the arxiv where we study the impact of data coming from the proton-lead run at the LHC. The full article is available at

Understanding the behaviour of partons inside a nucleus is crucial for the correct interpretation of experimental results in high energy collisions when at least one of the initial particles colliding is a nucleus. The LHC proton-lead run posses a high sensitivity to formerly unexplored kinematical regions and therefore will play a fundamental role in the determination of the nuclear parton densities, in particular for the gluon.

Obtaining the information from the data should be done in principle through the global analysis of all (new and old) published data. Unfortunately such a procedure is quite time consuming and expediency-lacking.

In our work we bypass this obstacle by means of a reweighting technique (Bayesian reweighting) and analyze the potential constraining power of the LHC proton-lead run data. The most relevant result is a public code that allows any user to appraise the impact on the EPS09 nuclear parton densities of any new data set. The code can be downloaded from:

By using this code we estimate a reduction of the nuclear gluon uncertainty up to a factor of two using some proton-nucleus observables as examples of the potentialities of these data