WGRAD : A parton level Monte Carlo program for the calculation of electroweak radiative corrections to W boson production at hadron colliders

Authors : Ulrich Baur, Stephane Keller, Doreen Wackeroth

The source code can be downloaded here : wgrad.tar.gz


Here is a brief WGRAD - manual :

1) You should have received 3 files: WGRAD.F, COMMON.INC and CONFIG.INC. That's the source code and two include files (common blocks). WGRAD.COM is a sample com file. The program calculates W production at O(alpha) including soft, virtual + hard one photon corrections from the initial and final state. Weak corrections are also taken into account. Initial and final state interference terms are correctly incorporated. The program can also be used to calculate W production at the Born level, or just the radiative corrections. In its next-to-leading mode, it evaluates the 2->2 and 2->3 (real photon radiation) parts. Both parts depend on cutoff parameters d_s and d_theta, but the sum of the differential cross sections is independent of these parameters. You can select electron or muon final states, and one can calculate initial or final state corrections, or the full schmu. For final state radiation, one can also select a mode where the program calculates W production in the approximation used by Berends + Kleiss.

2) You will be prompted for input for root s etc.... The program interfaces with PDFLIB and HBOOK. Notation as in V4.xx and higher of PDFLIB. For the collinear cutoff d_theta you have to choose a value smaller than the soft cutoff d_s. In the current version, d_s=0.01 and d_theta=0.001, but this can be changed easily. Cuts and detector related effects are included. You can either select no cuts/smearing and no recombination of electrons+photons, or smearing only, or smearing plus recombination. Look into CUTS to find out how we have done this. In the muon case, the recombination is replaced by an explicit isolation cut.

3) In the electron case with recombination, you can select whether you want the cross section computed with the collinear part of the integration performed explicitly, or numerically. The numerical method gives less smooth histograms. There is also a flag for the scale Q: you can chose mu=Q=M_W, 0.5*M_W or 2*M_W.

4) The four momenta of the final state particles (without smearing) and the event weights are stored in two common blocks: common/pnosmear_two/ptwo,wt2 and common/pnosmear_three/pthree,wt3 ptwo(i,j) (pthree(i,j)) for the 2->2 and 2->3 part. i=1,...4; the 4th component is the energy, j=1,2 (j=1,2,3) labels the final state particles, with j=3 for the photon. j=1: charged lepton, j=2: neutrino. wt2 and wt3 are the event weights. All cross sections are in nanobarns (!!).

5) Histograms are defined in SETUP_GRAPHS and filled in GRAPHS* and SETUP_PAW. A PAW file is dumped there as well. Data files suitable for plotting with topdrawer etc. are created through LISDAT at the end of the main routine.

6) Presently the programs run on VAX/VMS, Alphas and SUNS. It probably runs on Silly Graphs as well. With very minor modifications it runs on LINUX systems.

7) To compile the program use `for/extend_source/opt=(level=4,tune=host, unroll=16)' for best results on the Alpha or VAX and ` -O4 -extend_source ' for UNIX.

8) Link WGRAD, PDFLIB, PACKLIB, MATHLIB and KERNLIB to get the executable.

9) Our calculation fully takes into account the W decay into leptons, with all correlations and finite W width effects incorporated. ALL standard model graphs are included. The program calculates the EWK corrections in the MSbar scheme. There is a switch to switch over to the DIS scheme. If you need to know, drop me a line. Also: we currently use a constant W width. There is switch built in so that you can easily switch over to an energy dependent width.

10) To model the collinear (almost) singularity correctly, you need to generate a sufficient number of events (all our events are weighted!!). The smaller d_s and d_theta, the more events you need. d_s and d_theta must be smaller than 0.1. Above, the soft/collinear approximation which was used to calculate the soft+virtual corrections breaks down. A summary of the run parameters and cross section results is stored in the file WZ.RES which is created during the run.

11) The integration is done using RENO, a Madison variation of VEGAS.

12) Presently the QCD k-factor has been set = 1.

13) If you have questions: Please do not hesitate to contact Doreen Wackeroth.

E-mail: dw24 at buffalo.edu

Back to my Homepage