*** NEW April 2007 ***
Pomwig 2.0 beta now includes the H1 2006 Diffractive pdfs (H1 Collaboration Eur.Phys.J.C48:715-748,2006, hep-ex/0606004).
A tar archive, including all the old pomwig files and the new H1 pdfs, is here
H1 provide two diffractive pdfs, fit A and fit B. These fits have different pomeron and reggeon fluxes as well as different pomeron pdfs. The reggeon pdfs are the Owens pion pdfs (Phys.Rev.D30:943,1984) as in the previous 1997 fits, and previous versions of POMWIG. POMWIG must still therefore be interfaced to pdflib if you want to use the reggeon.
The NSTRU parameter is used to access the different pdfs:
NSTRU = 9,10 and 11 will give the H1 1997 pomeron, reggeon and user defined structure functions as in Pomwig 1.x.
For the above pdfs, everything will function as in Pomwig 1.x. So any main programs you have for use with previous versions will function as before.
The new pdfs are accessed as follows :
NSTRU = 12 : H1 2006 Fit A Pomeron
NSTRU = 13 : H1 2006 Fit A Reggeon
NSTRU = 14 : H1 2006 Fit B Pomeron
NSTRU = 15 : H1 2006 Fit B Reggeon
Note that you have to initialise the H1 routine qcd_2006.f routine in a similar way to the old h1qcd.f routine, using the ifit variable. An example main program is included in the tar file (pomwigmain.f ).
The initialisation (which we recommend you place in the main program in HWABEG) is as follows :
SUBROUTINE HWABEG
C USER'S ROUTINE FOR INITIALIZATION
INCLUDE 'HERWIG65.INC'
double precision XPQ(-6:6),X,Q2
double precision F2(2),FL(2),C2(2),CL(2)
integer ifit
C INITIALISE H1 POMERON
Q2=75
X=0.1
WRITE(*,*) 'INITIALISING H1 POMERON STRUCTURE'
IF (NSTRU.EQ.9) THEN
C Select your own ifit variable as before here - 1,2 and 3 for NLO 1997, 4,5 and 6 for LO 1997
ifit=5
call QCD_1994(X,Q2,XPQ,ifit)
C 2006 H1 pdfs
ELSEIF (NSTRU.EQ.12) THEN
ifit=1
call QCD_2006(X,Q2,ifit,XPQ,f2,fl,c2,cl)
ELSEIF (NSTRU.EQ.14) THEN
ifit=2
call QCD_2006(X,Q2,ifit,XPQ,f2,fl,c2,cl)
ENDIF
END
C---------------------------------------------------------------------
The example main program provided generates double pomeron -> Higgs events at LHC, Higgs mass 115 GeV, using H1 2006 Pomeron Fit A.
The generated cross section should be ~ 2.46 E-02 pb
*** NEW August 2005 ***
A new beta version of pomwig is now available. The only change is an option to treat the pomeron remnant differently. The Pomwig default is to treat the remnants of both the pomeron and reggeon in the same way that Herwig treats the photon remnant. This means that the valence partons in the pomeron and reggeon are defined to be a q - qbar pair. If a gluon from the pomeron enters the hard subprocess, for example, Herwig will backward evolve the gluon by parton showering until it produces a q - qbar pair.
The alternative is to define the valence partons in the pomeron to be gluons (which would seem a more natural choice). In this case, if a gluon from the pomeron enters the hard subprocess, there will tend to be less parton showering and therefore less remnant activity.
To try this new treatment, just replace the Herwig routines hwsspc.f and hwsval.f. Pomwig will then automatically treat the valence partons in the pomeron (or reggeon) as gluons.
Note that it is probably more sensible to generate reggeons in the standard way, since the valence partons might be more naturally assumed to be quarks (since the reggeon has a pion structure function).
We will probably introduce a more elegant way of selecting the remnant treatment in a future release.
As an example, we have noticed a significant difference in the simulation of double pomeron dijet results at the Tevatron, where the partons from the pomerons entering the hard subprocess are predominantly gluons.
*** NEW July 2004 ***
After many requests, we have modified the HWEGAM routine to change the outgoing electrons and positrons in the event record to protons and anti-protons respectively. This is implemented in the new 1.2 release, which also uses the NSTRU = 9,10 and 11 settings for pomeron, reggeon and user-defined structure functions in order to be compatible with HERWIG 6.5 - this will also function correctly with previous HERWIG versions.
Pomwig v1.2 is available for download here
*** NEW NOVEMBER 21 2002 ***
If you are using POMWIG with HERWIG 6.5 :
HERWIG 6.5 uses NSTRU = 6,7,8 internally, which means that POMWIG has had
to move to NSTRU = 9,10 and 11 for pomeron, reggeon and user-defined structure
functions respectively. Everything else will function as normal. Use the 1.2 version above if you are using HERWIG 6.5.
For users of Herwig 6.4 and earlier, the standard POMWIG v1.1 source code may
still be used, using NSTRU = 6,7,8.
Pomwig is a modified version of Herwig which is capable of generating diffractive interactions. All standard Herwig hard subprocesses are available in proton - pomeron, photon - pomeron and pomeron - pomeron collisions. Reggeons are also available.
A short POMWIG manual is available.
If you use Pomwig, please refer to it as :
B. E. Cox and J. R. Forshaw, Comput.Phys.Commun.144:104-110,2002
The POMWIG v1.1 source code can be obtained here. It consists of replacement routines for the HERWIG 5.9 routines HWSFUN and HWEGAM , plus the new routiunes H1QCD and FLUX, and POMSTR. The routines also function correctly with HERWIG 6.1. For details of installing and running POMWIG, please refer to the manual.
POMSTR.F is a new addition for v1.1, not described in hep-ph/0010303. It allows for the use of user defined structure functions. The default is x(1-x) for the up, down, anti - up, anti-down and gluon, and zero for all others. Call it with NSTRU=8. A flux for the user defined structure functions has been added to FLUX.F. The default is to use the same parameters as for the H1 pomeron structure function, with the normalisation coefficient in the flux set to one.
BUG FIXES IN VERSION 1.1
A bug in hwegam.f which affected only double pomeron collisions has been fixed
BUG FIXES IN VERSION 1.01
A bug in the reggeon code in hwegam.f has been fixed.
A bug in the H1 pomeron pdf call which resulted
in the structure function always being called with a fixed scale has been
fixed.
Please send any comments or questions to
coxb@mail.desy.de