KA-TP-05-2013MAN/HEP/2013/03MCnet-13-02 Multiple Partonic Interactions in Herwig++Presented by Andrzej Siódmok at 42. International Symposium on Multiparticle Dynamics (ISMD 2012), Kielce, Poland

Ka-Tp-05-2013 Man/hep/2013/03 MCnet-13-02 Multiple Partonic Interactions in Herwig++thanks: Presented by Andrzej Siódmok at 42. International Symposium on Multiparticle Dynamics (ISMD 2012), Kielce, Poland

Stefan Gieseke, Christian Röhr
Institut für Theoretische Physik, Karlsruhe Institute of Technology (KIT), Karlsruhe, Germany
Andrzej Siódmok
Consortium for Fundamental Physics, School of Physics and Astronomy,
The University of Manchester, Manchester, U.K.

We review the implementation of a model for multiple partonic interactions in Herwig++. Moreover, we show how recent studies on the colour structure of events in Herwig++ led to a significant improvement in the description of soft inclusive observables in pp interactions at the LHC.

PACS numbers: 11.80.La, 12.38.Lg

1 Introduction

Multiple partonic interaction (MPI) models are crucial for a successful description of the underlying event (UE) in hard hadronic collisions and minimum-bias (MB) data from the Tevatron and the Large Hadron Collider (LHC). The modern Monte Carlo event generators Herwig++ [1], Pythia [2, 3] and Sherpa [4] all include an MPI model in order to simulate the underlying event. In this short note we summarize the modelling of multiple parton interactions in the Herwig++ event generator, present its recent developments and show a comparison of the improved model to LHC data.

2 MPI model in Herwig++

The modelling of underlying events in Herwig++ is based on the fact that at high enough energies the hard inclusive cross section for dijet production,


will eventually exceed the total cross section [5]. This leads to the interpretation that in fact the inclusive cross section counts not only single hard events but all hard events that occur in parallel during the very same hadron-hadron collision.

2.1 Eikonal model

With the eikonal assumption that at fixed impact parameter, , individual scatterings are independent111Of course some basic correlations are incorporated into the model later on, for example by requirement of momentum conservation. and that the distribution of partons in hadrons factorizes with respect to the and dependence, we can express the average number of hard interactions in a hadron-hadron collision as:


where the function describes the spatial overlap of the two colliding hadrons (protons) as a function of the impact parameter . In both Fortran Herwig [6] with a Jimmy plug-in [7] and Herwig++, the overlap function is modelled according to the electromagnetic form factor [8] and the parameter interpreted as the inverse radius of the proton. Since the spatial parton distribution (the colour distribution) is assumed to be similar to the distribution of electric charge, but not necessarily identical, is treated as a free parameter. This model allows for the simulation of multiple interactions with perturbative scatters which have .

The extension to soft scatterings (with ) is kept as simple as possible. The additional soft contribution to the inclusive cross section is also eikonalized, such that we can also calculate an average number of soft scatters from the resulting and an overlap function for the soft scattering centres. The functional form is assumed to be the same as for the hard scatters, but we allow for a different inverse radius, . The consistency of this model with unitarity is given by fixing the two additional parameters and from two constraints. First, we can calculate the total cross section from the eikonal model and fix it to be consistent with the Donnachie-Landshoff parametrization [9]. In addition, using the optical theorem, we can calculate the –slope parameter from the eikonal model and fix it to a reasonable parametrization. This model and its extension to soft scatterings were implemented in Herwig++ [5, 10] and proved to be capable of describing the whole spectrum of UE data from the Tevatron [11, 12] (including its minimum bias part). However, the MPI model is too simple to describe the first MB data from ATLAS [13, 14] (see red line in Fig. 1).

Figure 1: Comparison of Herwig 2.4.2 without CR and Herwig 2.5 with PCR to ATLAS minimum-bias distributions at TeV.

A tuning of the parameters of the model [15] did not improve the description. This observation led to new developments of the MPI model to include non-perturbative colour reconnections (CR) [16, 17].

2.2 Colour reconnection

The colour structure of multiple interactions can cause non-trivial changes to the colour topology of the colliding system as a whole, with potentially major consequences for the particle multiplicity in the final state. The colour connections between partons define colour singlet objects known as clusters (see Fig. 2).

Figure 2: Formation of clusters, which we represent by ovals here. Colour lines are dashed. The right diagram shows a possible colour-reconnected state: the partons of the clusters and are arranged in new clusters, and .

The cluster hadronization model [18], which is used in Herwig++, generates clusters which are used as the starting point for the generation of hadrons via cluster decays. The idea of CR is based on colour preconfinement [19], which implies that parton jets emerging from different partonic interactions are colour-connected (clustered) if they are located closely in phase space. As the MPI model does not take that into account, those colour connections have to be adapted afterwards by means of a CR procedure pictured in Fig. 2. In our CR model we define the distance between two partons to be small when their invariant mass (cluster mass) is small. Therefore, the aim of the CR model is to reduce the colour length , where is the number of clusters in an event and is the invariant mass of cluster . Based on this physical motivation, there are two colour reconnection models implemented in Herwig++. The first CR model, so-called plain CR (PCR), has been included in Herwig++ as of version 2.5 [20]. This model iterates over all cluster pairs in a random order. Whenever a swap of colours is preferable, i.e. when the new cluster masses are smaller, this is done with a given probability (the only parameter of the model). The PCR model has been shown to give the desired results (see blue line in Fig. 1). However, as the clusters are presented to the model only in a given sequence, it is hard to assess which clusters are affected and to what extent the sequence is physically relevant. Therefore, we implemented another model, the statistical CR model (SCR) [17], which adopts the Metropolis algorithm to reduce . This allowed us to systematically study the consequences of the introduction of CR, the results of which are presented in details in [17, 21].

3 Tuning and Results

Initially we were primarily aiming at an improved description of MB data, therefore we started by tuning the PCR model to ATLAS MB data222Since currently there is no model for soft diffractive physics in Herwig++, we use the diffraction-reduced ATLAS MB measurement with an additional cut on the number of charged particles, Nch . (results are shown in Fig. 1). The next important question was whether the new model is able to describe the UE data sets at different collider energies. A dedicated tuning procedure [22] resulted in energy-extrapolated tunes UE-EE, for both PCR and SCR, in which all parameters are fixed except for , which varies with energy. These tunes allow us to describe data at different energies [11, 23] (see Fig. 3) with a simple parametrization of the energy dependence. This was an important step towards the understanding of the energy dependence of the model.

Figure 3: ATLAS data at GeV (1st column), CDF data at GeV (2nd column) and ATLAS data at TeV (3rd column), showing the multiplicity density of the charged particles in the “transverse” area as a function of .

Finally, we show the results of the model for energy flow as a function of for minimum-bias events at and 7 TeV (Fig. 4) presented by the CMS collaboration during this workshop and also published in [25]. This observable was not available during the preparation of the tunes and these very good results can therefore be treated as a prediction of the model.

Figure 4: Energy flow as a function of for MB events at and 7 TeV.

The new model is implemented and available with the recent tunes in Herwig++ version 2.6 [24].

4 Summary

We have summarized the multiple partonic interaction models in Herwig++ and expanded on the motivation and modelling of colour reconnection. Furthermore, we have shown that (sufficiently diffraction-suppressed) MB data from the LHC and underlying-event observables are well described by the present model which makes it useful for the LHC collaborations.

5 Acknowledgements

We thank the organizers for the very pleasant and fruitful workshop. This work has been supported by the Helmholtz Alliance ”Physics at the Terascale“ and by the Lancaster-Manchester-Sheffield Consortium for Fundamental Physics under STFC grant ST/J000418/1. We wish to acknowledge Dermot Moran for his critical reading of this proceedings.


  • [1] M. Bähr et al., Eur. Phys. J. C 58 (2008) 639
  • [2] T. Sjöstrand, S. Mrenna and P. Z. Skands, JHEP 0605 (2006) 026
  • [3] T. Sjöstrand, S. Mrenna, P. Z. Skands, Comput. Phys. Commun. 178 (2008)
  • [4] T. Gleisberg, S. Hoeche, and F. Krauss et al. JHEP 0902 (2009) 007
  • [5] M. Bähr, J. M. Butterworth and M. H. Seymour, JHEP 0901 (2009) 065
  • [6] G. Corcella, I. G. Knowles, G. Marchesini, S. Moretti, K. Odagiri, P. Richardson, M. H. Seymour and B. R. Webber, JHEP 0101 (2001) 010
  • [7] J. M. Butterworth, J. R. Forshaw and M. H. Seymour, Z. Phys. C 72 (1996)
  • [8] I. Borozan and M. H. Seymour, JHEP 0209 (2002) 015
  • [9] A. Donnachie and P. V. Landshoff, Phys. Lett. B 296 (1992) 227
  • [10] M. Bähr, J. M. Butterworth, S. Gieseke and M. H. Seymour, arXiv:0905.4671.
  • [11] T. Affolder et al. [CDF Collaboration], Phys. Rev. D 65 (2002) 092002.
  • [12] D. Acosta et al. [CDF Collaboration], Phys. Rev. D 70 (2004) 072002
  • [13] G. Aad et al. [ATLAS Collaboration], Phys. Lett. B 688 (2010) 21
  • [14] G. Aad et al. [ATLAS Collaboration], New J. Phys. 13 (2011) 053033
  • [15] P. Bartalini, E. L. Berger, B. Blok and G. Calucci, et al., arXiv:1111.0469.
  • [16] S. Gieseke, C. A. Röhr and A. Siódmok, arXiv:1110.2675 [hep-ph].
  • [17] S. Gieseke, C. Röhr and A. Siódmok, Eur. Phys. J. C 72 (2012) 2225
  • [18] B.R. Webber, Nuclear Phys. B 238 (1984) 492
  • [19] D. Amati, and G. Veneziano Phys. Lett. B83 (1979) 87
  • [20] S. Gieseke, D. Grellscheid, and K. Hamilton et al., arXiv:1102.1672 [hep-ph].
  • [21] S. Gieseke, C. Röhr and A. Siódmok, arXiv:1206.2205 [hep-ph].
  • [22] S. Gieseke, C. A. Röhr and A. Siódmok, DESY-PROC-2012-03.
  • [23] G. Aad et al. [Atlas Collaboration], Phys. Rev. D 83 (2011) 112001
  • [24] K. Arnold, L. d’Errico, and S. Gieseke et al., arXiv:1205.4902 [hep-ph].
  • [25] S. Chatrchyan et al. [CMS Collaboration], JHEP 1111 (2011) 148
Comments 0
Request Comment
You are adding the first comment!
How to quickly get a good reply:
  • Give credit where it’s due by listing out the positive aspects of a paper before getting into which changes should be made.
  • Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
  • Your comment should inspire ideas to flow and help the author improves the paper.

The better we are at sharing our knowledge with each other, the faster we move forward.
The feedback must be of minimum 40 characters and the title a minimum of 5 characters
Add comment
Loading ...
This is a comment super asjknd jkasnjk adsnkj
The feedback must be of minumum 40 characters
The feedback must be of minumum 40 characters

You are asking your first question!
How to quickly get a good answer:
  • Keep your question short and to the point
  • Check for grammar or spelling errors.
  • Phrase it like a question
Test description