Progress on Event Generation with PYTHIA 8
Progress on Event Generation with PYTHIA 8
Progress on Event Generation with PYTHIA 8
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
Collider Cross Talk<br />
CERN, 4 August 2011<br />
<str<strong>on</strong>g>Progress</str<strong>on</strong>g> <strong>on</strong> <strong>Event</strong> Generati<strong>on</strong><br />
<strong>with</strong> <strong>PYTHIA</strong> 8<br />
Torbjörn Sjöstrand<br />
Department of Astr<strong>on</strong>omy and Theoretical Physics<br />
Lund University<br />
Introducti<strong>on</strong> and Overview<br />
Improved Showers and ME matching<br />
MultiPart<strong>on</strong> Interacti<strong>on</strong>s<br />
BSM Physics<br />
Tunes and Comparis<strong>on</strong>s <strong>with</strong> LHC Data<br />
Summary and Outlook
Introducti<strong>on</strong> and Overview<br />
General-purpose event generators: <strong>PYTHIA</strong>, HERWIG, SHERPA<br />
<strong>PYTHIA</strong> size: ∼80,000 lines (Fortran in <strong>PYTHIA</strong> 6, C++ in <strong>PYTHIA</strong> 8)<br />
PDF<br />
ME<br />
MPI<br />
ISR<br />
FSR<br />
BR<br />
CR<br />
Hadr.<br />
Decays<br />
Unknown?
<strong>PYTHIA</strong> 8 ambiti<strong>on</strong><br />
• Meet experimental request for C++ code.<br />
• Housecleaning ⇒ more homogeneous.<br />
• More user-friendly (e.g. settings names).<br />
• Better match to software frameworks<br />
(e.g. card files).<br />
• More space for growth.<br />
• Better interfaces to external standards.<br />
Reality<br />
• Work begun autumn 2004.<br />
• 3 years at CERN ⇒ good progress.<br />
• First release autumn 2007.<br />
• Since then: slower progress,<br />
requests lagging behind.<br />
• Usage slowly taking off.<br />
Team members<br />
Stefan Ask<br />
Richard Corke<br />
Stephen Mrenna<br />
Peter Skands<br />
C<strong>on</strong>tributors include<br />
O. Alvestad, B. Bellenot,<br />
R. Brun, L. Carl<strong>on</strong>i,<br />
N. Desai, N. Hod,<br />
H. Hoeth, P. Ilten,<br />
T. Kasemets, M. Kirsanov,<br />
B. Lloyd, M. M<strong>on</strong>tull,<br />
A. Morsch, A. Naumann,<br />
S. Navin, P. Newman,<br />
S. Prestel, M. Sutt<strong>on</strong><br />
MSTW, CTEQ, H1: PDFs<br />
DELPHI, LHCb: D/B BRs<br />
+ several bug reports & fixes
Key differences between <strong>PYTHIA</strong> 6.4 and 8.1<br />
Old features definitely removed include, am<strong>on</strong>g others:<br />
• independent fragmentati<strong>on</strong><br />
• mass-ordered showers<br />
Features omitted so far include, am<strong>on</strong>g others:<br />
• ep, γp and γγ beam c<strong>on</strong>figurati<strong>on</strong>s<br />
• some processes, notably Technicolor<br />
New features, not found in 6.4:<br />
⋆ fully interleaved p ⊥-ordered MPI + ISR + FSR evoluti<strong>on</strong><br />
⋆ richer mix of underlying-event processes (γ, J/ψ, DY, . . . )<br />
⋆ possibility for two selected hard interacti<strong>on</strong>s in same event<br />
⋆ allow rescattering and x-dependent prot<strong>on</strong> size in MPI framework<br />
⋆ full hadr<strong>on</strong>–hadr<strong>on</strong> collisi<strong>on</strong> machinery for diffractive systems<br />
⋆ several new processes, <strong>with</strong>in and bey<strong>on</strong>d SM<br />
⋆ possibility to use <strong>on</strong>e PDF set for hard process and another for rest<br />
⋆ τ lept<strong>on</strong> polarizati<strong>on</strong> in producti<strong>on</strong> and decay<br />
⋆ updated decay data and LO PDF sets
Interleaved evoluti<strong>on</strong><br />
• Transverse-momentum-ordered part<strong>on</strong> showers for ISR and FSR<br />
• MPI also ordered in p ⊥<br />
Allows interleaved evoluti<strong>on</strong> for ISR, FSR and MPI<br />
dP<br />
dp ⊥<br />
=<br />
dPMPI<br />
× exp<br />
dp<br />
<br />
⊥<br />
−<br />
+ dPISR +<br />
dp⊥ <br />
dPFSR dp<br />
<br />
⊥<br />
dPMPI<br />
dp ′ +<br />
⊥<br />
dPISR dp ′ ⊥<br />
p⊥max<br />
p ⊥<br />
Ordered in decreasing p ⊥ using “Sudakov” trick.<br />
Corresp<strong>on</strong>ds to increasing “resoluti<strong>on</strong>”:<br />
smaller p ⊥ fill in details of basic picture set at larger p ⊥.<br />
Hybrid approach to shower recoils:<br />
• FSR is dipole: nearest colour-c<strong>on</strong>nected neighbour<br />
• ISR is traditi<strong>on</strong>al: whole hard-scattering system affected<br />
(since ISR dipole gives wr<strong>on</strong>g answer e.g. for p ⊥Z)<br />
+ dPFSR dp ′ <br />
dp<br />
⊥<br />
′ <br />
⊥
Shower matching to MEs: realistic hard default<br />
Aim: provide better default shower behaviour at large p⊥, to bridge gap between “power” and “wimpy” showers.<br />
dP / dp⊥ [GeV -1<br />
]<br />
10 -2<br />
10 -3<br />
10 -4<br />
10 -5<br />
10 -6<br />
10 -7<br />
(b)<br />
POWHEG<br />
Pythia Default (Power)<br />
Pythia Damp, k = 2<br />
Pythia Damp, k = 1<br />
Pythia Wimpy<br />
100 200 300 400 500 600 700 800 900 1000<br />
p⊥ [GeV]<br />
dP ISR<br />
dp 2 ⊥<br />
∝ 1<br />
p 2 ⊥<br />
k 2 M 2<br />
k 2 M 2 + p 2 ⊥<br />
tt producti<strong>on</strong><br />
M 2 = m 2 ⊥t<br />
= m 2 t + p2 ⊥t<br />
for coloured final state<br />
No dampening for uncoloured final state (W + W − , . . . , SUSY).<br />
R. Corke & TS, Eur. Phys. J. C69 (2010) 1
Shower matching to MEs: realistic QCD default<br />
Must avoid doublecounting for QCD jets:<br />
shower starting scale = p ⊥ of hard 2 → 2 process.<br />
Study how well the part<strong>on</strong> shower fills the phase space,<br />
as prelude to full matching to 2 → 3 real-emissi<strong>on</strong>:<br />
dσ / dp ⊥ [nb / GeV]<br />
10 2<br />
10 1<br />
10 0<br />
(a) p ⊥3<br />
PS<br />
ME<br />
10 -1<br />
0 10 20 30 40 50<br />
p⊥ [GeV]<br />
p min<br />
⊥3<br />
dσ / dp ⊥ [nb / GeV]<br />
10 2<br />
10 1<br />
10 0<br />
10 -1<br />
(b) p ⊥4<br />
PS<br />
ME<br />
10 -2<br />
0 10 20 30 40 50<br />
p⊥ [GeV]<br />
= 5.0 GeV, pmin<br />
⊥4 = 5.0 GeV, Rsep = 0.10<br />
dσ / dp ⊥ [nb / GeV]<br />
10 3<br />
10 2<br />
10 1<br />
10 0<br />
10 -1<br />
10 -2<br />
10 -3<br />
(c) p ⊥5<br />
PS<br />
ME<br />
10 -4<br />
0 10 20 30 40 50<br />
p⊥ [GeV]<br />
Obtain good qualitative agreement, best in soft and collinear regi<strong>on</strong>s,<br />
but large regi<strong>on</strong> of phase space well described, and <strong>on</strong>ly corners bad.<br />
No indicati<strong>on</strong> for needing a change in starting scale!<br />
R. Corke & TS, JHEP 03 (2011) 032
Shower matching to MEs: CKKW–L<br />
Leif Lönnblad and Stefan Prestel work to include CKKW–L matching<br />
• for jets, W/Z + jets, WW/WZ/ZZ + jets<br />
• rec<strong>on</strong>structing ME “history” as viewed by <strong>PYTHIA</strong> 8 shower<br />
• c<strong>on</strong>sistently taking into account effects of MPI<br />
(1/σ) dσ/dk ⊥ 1<br />
Deviati<strong>on</strong> [%]<br />
10 -7<br />
10 -8<br />
10 -9<br />
150<br />
100<br />
50<br />
0<br />
wME1PS<br />
wME2PS<br />
wME3PS<br />
wME4PS<br />
<strong>PYTHIA</strong>8 (no mi)<br />
ME 0 3PS (no mi)<br />
ME 1 3PS (no mi)<br />
ME 2 3PS (no mi)<br />
ME 3 3PS (no mi)<br />
wME3PS (no mi)<br />
-50<br />
0 20 40 60 80 100<br />
k⊥ 1 (GeV)<br />
Deviati<strong>on</strong> [%]<br />
200<br />
150<br />
100<br />
50<br />
0<br />
-50<br />
wME1PS<br />
wME2PS<br />
wME3PS<br />
wME4PS<br />
Article to appear any day now; code “so<strong>on</strong>” in <strong>PYTHIA</strong> 8;<br />
NLO matching natural extensi<strong>on</strong><br />
(1/σ) dσ/dk ⊥ 2<br />
10 -7<br />
10 -8<br />
10 -9<br />
10 -10<br />
<strong>PYTHIA</strong>8 (no mi)<br />
ME 0 3PS (no mi)<br />
ME 1 3PS (no mi)<br />
ME 2 3PS (no mi)<br />
ME 3 3PS (no mi)<br />
wME3PS (no mi)<br />
0 10 20 30 40 50 60 70<br />
k⊥ 2 (GeV)
Shower matching to MEs: POWHEG<br />
Standard Les Houches interface (LHA, LHEF) specifies startup scale SCALUP<br />
for showers, so “trivial” to interface any external program, including POWHEG.<br />
Problem: for ISR<br />
p 2 ⊥ = p2 ⊥evol − p4 ⊥evol<br />
p 2 ⊥evol,max<br />
i.e. p ⊥ decreases for θ ∗ > 90 ◦ but p ⊥evol m<strong>on</strong>ot<strong>on</strong>ously increasing.<br />
Soluti<strong>on</strong>: run “power” shower but kill emissi<strong>on</strong>s above the hardest <strong>on</strong>e,<br />
by POWHEG’s definiti<strong>on</strong>.<br />
(1/N) dN / dx<br />
1.2<br />
1<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
0<br />
(a)<br />
Factorisati<strong>on</strong> Scale<br />
Kinematical Limit + Veto<br />
0 0.2 0.4 0.6 0.8 1 1.2<br />
x = p ⊥ shower<br />
/ p ⊥ hard<br />
(1/N) dN / dx<br />
0.7<br />
0.6<br />
0.5<br />
0.4<br />
0.3<br />
0.2<br />
0.1<br />
0<br />
(b)<br />
Factorisati<strong>on</strong> Scale<br />
Kinematical Limit + Veto<br />
0 0.2 0.4 0.6 0.8 1 1.2<br />
x = p ⊥ shower<br />
Available for ISR-dominated, coming for QCD jets <strong>with</strong> FSR issues.<br />
/ p ⊥ hard
Shower matching to MEs: VINCIA<br />
Plug-in replacement to default <strong>PYTHIA</strong> showers; for now FSR <strong>on</strong>ly.<br />
Developed by Walter Giele, David Kosower, Peter Skands, . . .<br />
See talk by Juan Lopez-Villarejo <strong>on</strong> Wednesday 10 August!<br />
Shower matching to MEs: the rest<br />
As in <strong>PYTHIA</strong> 6, it is possible to use external ME generators<br />
<strong>with</strong> their own matching, e.g.<br />
• ALPGEN (<strong>with</strong> MLM)<br />
• MADGRAPH (<strong>with</strong> MLM)
Multipart<strong>on</strong> interacti<strong>on</strong>s key aspect<br />
of <strong>PYTHIA</strong> since > 20 years.<br />
Central to obtain agreement <strong>with</strong> data:<br />
Tune A, Professor, Perugia, . . .<br />
A sec<strong>on</strong>d hard process<br />
Before 8.1 no chance to select character of sec<strong>on</strong>d interacti<strong>on</strong>.<br />
Now free choice of first process (including LHA/LHEF)<br />
and sec<strong>on</strong>d process combined from list:<br />
• TwoJets (<strong>with</strong> TwoBJets as subsample)<br />
• Phot<strong>on</strong>AndJet, TwoPhot<strong>on</strong>s<br />
• Charm<strong>on</strong>ium, Bottom<strong>on</strong>ium (colour octet framework)<br />
• SingleGmZ, SingleW, GmZAndJet, WAndJet<br />
• TopPair, SingleTop<br />
Can be expanded am<strong>on</strong>g existing processes as need arises.<br />
By default same phase space cuts as for “first” hard process<br />
=⇒ sec<strong>on</strong>d can be harder than first.<br />
However, possible to set ˆm and pˆ ⊥ range separately.
Multipart<strong>on</strong> interacti<strong>on</strong>s<br />
Regularise cross secti<strong>on</strong> <strong>with</strong> p ⊥0 as free parameter<br />
dˆσ<br />
dp 2 ⊥<br />
<strong>with</strong> energy dependence<br />
∝ α2 s(p 2 ⊥ )<br />
p 4 ⊥<br />
p ⊥0(E CM) = p ref<br />
⊥0 ×<br />
→ α2 s(p 2 ⊥0 + p2 ⊥ )<br />
(p 2 ⊥0 + p2 ⊥ )2<br />
<br />
ECM<br />
Eref ǫ CM<br />
Matter profile in impact-parameter space<br />
gives time-integrated overlap which determines level of activity:<br />
simple Gaussian or more peaked variants<br />
ISR and MPI compete for beam momentum → PDF rescaling<br />
+ flavour effects (valence, qq pair compani<strong>on</strong>s, . . . )<br />
+ correlated primordial k ⊥ and colour in beam remnant<br />
Many part<strong>on</strong>s produced close in space–time ⇒ colour rearrangement;<br />
reducti<strong>on</strong> of total string length ⇒ steeper 〈p ⊥〉(n ch)
Often<br />
assume<br />
that<br />
MPI =<br />
Rescattering<br />
. . . but<br />
should<br />
also<br />
include<br />
Same order in αs, ∼ same propagators, but<br />
• <strong>on</strong>e PDF weight less ⇒ smaller σ<br />
• <strong>on</strong>e jet less ⇒ QCD radiati<strong>on</strong> background 2 → 3 larger than 2 → 4<br />
⇒ will be tough to find direct evidence.<br />
Rescattering grows <strong>with</strong> number of “previous” scatterings:<br />
Tevatr<strong>on</strong> LHC<br />
Min Bias QCD Jets Min Bias QCD Jets<br />
Normal scattering 2.81 5.09 5.19 12.19<br />
Single rescatterings 0.41 1.32 1.03 4.10<br />
Double rescatterings 0.01 0.04 0.03 0.15<br />
R. Corke & TS, JHEP 01 (2010) 035
An x-dependent prot<strong>on</strong> size<br />
Normally assume that PDFs factorize in l<strong>on</strong>gitudinal and transverse space:<br />
f(x, r) = f(x) ρ(r)<br />
In c<strong>on</strong>tradicti<strong>on</strong> <strong>with</strong><br />
• intuitive picture of part<strong>on</strong>s spreading out by cascade to lower x<br />
• formally BFKL, Balitsky-JIMWLK, Colour Glass C<strong>on</strong>densate, . . .<br />
• Mueller’s dipole cascade (Lund program: DIPSY; study by Avsar)<br />
• Froissart-Martin σ tot ∝ ln 2 s by Gribov theory related to rp ∝ ln(1/x)<br />
• generalized part<strong>on</strong> distributi<strong>on</strong>s, . . .<br />
For now address <strong>on</strong>ly inelastic n<strong>on</strong>diffrative events,<br />
σtot = σel + σSD + σDD + σND, <strong>with</strong> ansatz:<br />
ρ(r, x) ∝ 1<br />
a 3 (x) exp<br />
<br />
− r2<br />
a 2 (x)<br />
<br />
<strong>with</strong> a(x) = a0<br />
<br />
1 + a1ln 1<br />
x<br />
a1 ≈ 0.15 tuned to rise of σ ND in D<strong>on</strong>nachie & Landshoff + Schuler & TS<br />
a0 tuned to value of σ ND, given PDF, p ⊥0, . . .<br />
R. Corke & TS, JHEP 05 (2011) 009
C<strong>on</strong>voluti<strong>on</strong> of two incoming prot<strong>on</strong>s gives impact parameter shape<br />
Õ(b; x1, x2) = 1<br />
π<br />
1<br />
a 2 (x1) + a 2 (x2) exp<br />
<br />
b<br />
−<br />
2<br />
a2 (x1) + a2 (x2)<br />
Define n(b) as average number of interacti<strong>on</strong>s for passage at b<br />
n(b) = <br />
i,j<br />
<br />
dx1dx2dp 2 ⊥ f i(x1, p 2 ⊥ ) f j(x2, p 2 ⊥<br />
such that σ hard =<br />
<br />
n(b)d 2 b<br />
σND = Pintd 2 b =<br />
<br />
which gives a 〈b2 〉 for σhard and σND: √〈b 2 〉 [fm]<br />
0.85<br />
0.8<br />
0.75<br />
0.7<br />
0.65<br />
0.6<br />
0.55<br />
10 2<br />
a1 = 0.00<br />
a1 = 0.15<br />
a1 = 1.00<br />
10 3<br />
(a)<br />
E CM [GeV]<br />
10 4<br />
<br />
10 5<br />
√〈b 2 〉 eik [fm]<br />
1.2<br />
1.1<br />
1<br />
0.9<br />
0.8<br />
0.7<br />
10 2<br />
a1 = 0.00<br />
a1 = 0.15<br />
a1 = 1.00<br />
) dˆσij<br />
dp 2 ⊥<br />
<br />
<br />
<br />
<br />
<br />
reg<br />
<br />
1 − e −n(b)<br />
d 2 b<br />
10 3<br />
(b)<br />
E CM [GeV]<br />
10 4<br />
<br />
Õ(b; x1, x2)<br />
10 5
C<strong>on</strong>sequence: collisi<strong>on</strong>s at large x will have to happen at small b,<br />
and hence further large-to-medium-x MPIs are enhanced,<br />
while low-x part<strong>on</strong>s are so spread out that it plays less role.<br />
norm<br />
(1 / N) dN / dbMPI (1 / N) dN / dN MPI<br />
1.8<br />
1.6<br />
1.4<br />
1.2<br />
1<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
0<br />
0.1<br />
0.08<br />
0.06<br />
0.04<br />
0.02<br />
0<br />
(a)<br />
0 0.5 1 1.5 2 2.5<br />
norm<br />
bMPI (b) Z 0<br />
SG<br />
DY<br />
Z 0<br />
Z’<br />
0 5 10 15 20 25<br />
N MPI<br />
SG<br />
Log<br />
(1 / N) dN / dN MPI<br />
(1 / N) dN / dN MPI<br />
0.1<br />
0.08<br />
0.06<br />
0.04<br />
0.02<br />
0<br />
0.12<br />
0.1<br />
0.08<br />
0.06<br />
0.04<br />
0.02<br />
0<br />
(a) DY<br />
0 5 10 15 20 25<br />
N MPI<br />
(c) Z’<br />
SG<br />
Log<br />
0 5 10 15 20 25<br />
N MPI<br />
SG<br />
Log
Diffracti<strong>on</strong><br />
Ingelman-Schlein: Pomer<strong>on</strong> as hadr<strong>on</strong> <strong>with</strong> part<strong>on</strong>ic c<strong>on</strong>tent<br />
Diffractive event = (Pomer<strong>on</strong> flux) × (IPp collisi<strong>on</strong>)<br />
p<br />
p<br />
IP<br />
p<br />
Used e.g. in<br />
POMPYT<br />
POMWIG<br />
PHOJET<br />
1) σ SD and σ DD taken from existing parametrizati<strong>on</strong> or set by user.<br />
2) Shape of Pomer<strong>on</strong> distributi<strong>on</strong> inside a prot<strong>on</strong>, f IP/p (x IP, t)<br />
gives diffractive mass spectrum and scattering p ⊥ of prot<strong>on</strong>.<br />
3) At low masses retain old framework, <strong>with</strong> l<strong>on</strong>gitudinal string(s).<br />
Above 10 GeV begin smooth transiti<strong>on</strong> to IPp handled <strong>with</strong> full pp<br />
machinery: multiple interacti<strong>on</strong>s, part<strong>on</strong> showers, beam remnants, . . . .<br />
4) Choice between 5 Pomer<strong>on</strong> PDFs.<br />
Free parameter σ IPp needed to fix 〈n interacti<strong>on</strong>s〉 = σ jet/σ IPp.<br />
5) Framework needs testing and tuning, e.g. of σ IPp.
Beate Heinemann, MB/UE Working Group (also Sparsh Navin)
BSM Physics 1: R-parity violati<strong>on</strong><br />
Encountered in R-parity violating SUSY decays ˜χ 0 1 → uds,<br />
or when 2 valence quarks kicked out of prot<strong>on</strong> beam<br />
d (g)<br />
juncti<strong>on</strong><br />
lab frame d (g)<br />
rest frame<br />
s (b)<br />
d<br />
s<br />
q4<br />
q5<br />
J<br />
q4<br />
q5<br />
q3<br />
flavour space<br />
x<br />
u (r)<br />
q3 q2 q2 qq1 qq1 u<br />
z<br />
120 ◦<br />
s (b)<br />
J<br />
120 ◦<br />
120 ◦<br />
More complicated<br />
(but ≈solved) <strong>with</strong><br />
glu<strong>on</strong> emissi<strong>on</strong> and<br />
massive quarks<br />
P. Skands & TS, Nucl. Phys. B659 (2003) 243<br />
u (r)
BSM Physics 2: R-hadr<strong>on</strong>s<br />
What if coloured (SUSY) particle like ˜g or˜t1 is l<strong>on</strong>g-lived?<br />
⋆ Formati<strong>on</strong> of R-hadr<strong>on</strong>s<br />
˜gqq ˜t1q “mes<strong>on</strong>s”<br />
˜gqqq ˜t1qq “bary<strong>on</strong>s”<br />
˜gg “glueballs”<br />
⋆ C<strong>on</strong>versi<strong>on</strong> between R-hadr<strong>on</strong>s<br />
by “low-energy” interacti<strong>on</strong>s <strong>with</strong> matter:<br />
˜gud + p → ˜guud + π + irreversible<br />
⋆ Displaced vertices if finite lifetime, or else<br />
⋆ punch-through: σ ≈ σ had but<br />
∆E < ∼ 1 GeV ≪ E kin,R<br />
A.C. Kraan, Eur. Phys. J. C37 (2004) 91;<br />
M. Fairbairn et al., Phys. Rep. 438 (2007) 1<br />
CMS, arXiv:1101.1645<br />
Partly event generati<strong>on</strong>, partly detector simulati<strong>on</strong>.<br />
Public add-<strong>on</strong> in <strong>PYTHIA</strong> 6, now integrated part of <strong>PYTHIA</strong> 8.<br />
Can also be applied to n<strong>on</strong>-SUSY l<strong>on</strong>g-lived “hadr<strong>on</strong>s”.
BSM Physics 3: Hidden Valley (Secluded Sector)<br />
What if new gauge groups at low energy scales, hidden by<br />
potential barrier or weak couplings? (M. Strassler & K. Zurek, . . . )<br />
Complete framework implemented in <strong>PYTHIA</strong>:<br />
⋆ New gauge group either Abelian U(1) or n<strong>on</strong>-Abelian SU(N)<br />
⋆ 3 alternative producti<strong>on</strong> mechanisms<br />
1) massive Z ′ : qq → Z ′ → qvqv<br />
2) kinetic mixing: qq → γ → γv → qvqv<br />
3) massive Fv charged under both SM and hidden group<br />
⋆ Interleaved shower in<br />
QCD, QED and HV sectors:<br />
add qv → qvγv (and Fv)<br />
or qv → qvgv, gv → gvgv,<br />
which gives recoil effects<br />
also in visible sector<br />
L. Carl<strong>on</strong>i & TS, JHEP 09 (2010) 105;<br />
L. Carl<strong>on</strong>i, J. Rathsman & TS, JHEP 04 (2011) 091
⋆ Hidden Valley particles may remain invisible, or . . .<br />
⋆ Broken U(1): γv acquire mass, radiated γvs decay back<br />
γv → γ → ff <strong>with</strong> BRs as phot<strong>on</strong> (⇒ lept<strong>on</strong> pairs!)<br />
⋆ SU(N): hadr<strong>on</strong>izati<strong>on</strong> in hidden sector, <strong>with</strong> full string fragmentati<strong>on</strong>,<br />
permitting up to 8 different qv flavours and 64 qvqv mes<strong>on</strong>s,<br />
but for now assumed degenerate in mass, so <strong>on</strong>ly distinguish<br />
– off-diag<strong>on</strong>al, flavour-charged, stable & invisible<br />
– diag<strong>on</strong>al, can decay back qvqv → ff<br />
Even when tuned to same average activity, hope to separate U(1) and SU(N):<br />
#(N v )<br />
0.4<br />
0.35<br />
0.3<br />
0.25<br />
0.2<br />
0.15<br />
0.1<br />
0.05<br />
0<br />
0 2 4 6 8 10<br />
N v<br />
Ab<br />
n<strong>on</strong>-Ab.<br />
#(cos θ)<br />
0.8<br />
0.7<br />
0.6<br />
0.5<br />
0.4<br />
0.3<br />
0.2<br />
0.1<br />
0<br />
Ab.<br />
n<strong>on</strong>-Ab.<br />
0 0.5 1 1.5 2 2.5 3<br />
cos θ
Tuning prospects<br />
Tuning to e + e − closely related to p ⊥-ordered <strong>PYTHIA</strong> 6.4;<br />
Rivet+Professor (H. Hoeth) ⇒ FSR & hadr<strong>on</strong>izati<strong>on</strong> OK (?)<br />
First tuning to MB data by P. Skands:<br />
⇒ MPI & colour rec<strong>on</strong>necti<strong>on</strong> OK (?)
But Rivet+Professor (H. Hoeth) shows it fails miserably for UE<br />
(Rick Field’s transverse flow as functi<strong>on</strong> of jet p ⊥):<br />
No universal tune MB + UE!<br />
Where did we go wr<strong>on</strong>g?
Final-state part<strong>on</strong> may have colour partner in the initial state.<br />
How to subdivide FSR and ISR in this kind of dipole?<br />
Large mass → large rapidity range for emissi<strong>on</strong>:<br />
In dipole rest frame<br />
(think rapidity space)<br />
m >> p⊥<br />
ISR<br />
m ∼ p⊥<br />
FSR<br />
m/2<br />
Soluti<strong>on</strong>: suppress final-state radiati<strong>on</strong> in double-counted regi<strong>on</strong><br />
p ⊥
Is a simultaneous MB/UE Tevatr<strong>on</strong> tune now possible?<br />
Tunes 2C and 2M d<strong>on</strong>e “by hand” (= using Rivet, but not Professor),<br />
using the CTEQ6L1 and MRST LO** PDF sets, respectively,<br />
to MB data (n ch, 〈p ⊥〉(n ch), . . . .<br />
Compare against Pro-Q20 and Perugia 0 (<strong>PYTHIA</strong> 6)<br />
〈N ch〉/dη dφ<br />
MC/data<br />
1<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
0<br />
1.4<br />
1.2<br />
1<br />
0.8<br />
0.6<br />
Transverse regi<strong>on</strong> charged particle density<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
CDF data<br />
<strong>PYTHIA</strong> 6.422 Pro-Q20<br />
<strong>PYTHIA</strong> 6.422 Perugia 0<br />
<strong>PYTHIA</strong> 8.142 Tune 2C<br />
<strong>PYTHIA</strong> 8.142 Tune 2M<br />
0 50 100 150 200 250 300 350 400<br />
pT(leading jet) / GeV<br />
<br />
<br />
<br />
〈∑ ptrack T 〉/dη dφ / GeV<br />
2<br />
1.5<br />
1<br />
0.5<br />
0<br />
1.4<br />
Transverse regi<strong>on</strong> charged ∑ p ⊥ density<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
CDF data<br />
<strong>PYTHIA</strong> 6.422 Pro-Q20<br />
<strong>PYTHIA</strong> 6.422 Perugia 0<br />
<strong>PYTHIA</strong> 8.142 Tune 2C<br />
<strong>PYTHIA</strong> 8.142 Tune 2M<br />
0 50 100 150 200 250 300 350 400<br />
pT(leading jet) / GeV<br />
Now generally comparably good descripti<strong>on</strong> to old tunes at the Tevatr<strong>on</strong>.<br />
MC/data<br />
1.2<br />
1<br />
0.8<br />
0.6
Tune 2C applied to LHC data does not do so good:<br />
1/N ev dN ev / dN ch<br />
<br />
10 2<br />
10 1<br />
10 0<br />
10 -1<br />
10 -2<br />
10 -3<br />
10 -4<br />
10 -5<br />
10 -6<br />
10 -7<br />
1.4<br />
1.2<br />
1.0<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
0.0<br />
ATLAS 7.00 TeV (x 100)<br />
Pythia 7.00 TeV (x 100)<br />
ATLAS 900 GeV<br />
Pythia 900 GeV<br />
0 10 20 30 40 50<br />
Nch 60 70 80 90 100<br />
ATLAS 7.00 TeV<br />
Pythia 7.00 TeV<br />
ATLAS 900 GeV<br />
Pythia 900 GeV<br />
0 2 4 6 8 10 12 14 16 18 20<br />
p T lead [GeV]<br />
Transverse<br />
regi<strong>on</strong><br />
1/N ev dN ev / dN ch<br />
[GeV]<br />
10 3<br />
10 2<br />
10 1<br />
10 0<br />
10 -1<br />
10 -2<br />
10 -3<br />
10 -4<br />
10 -5<br />
1.8<br />
1.6<br />
1.4<br />
1.2<br />
1<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
0<br />
ALICE 7.00 TeV (x 100)<br />
Pythia 7.00 TeV (x 100)<br />
ALICE 2.36 TeV (x 10)<br />
Pythia 2.36 TeV (x 10)<br />
ALICE 900 GeV<br />
Pythia 900 GeV<br />
0 10 20 30 40<br />
Nch 50 60 70 80<br />
ATLAS 7.00 TeV<br />
Pythia 7.00 TeV<br />
ATLAS 900 GeV (-0.5)<br />
Pythia 900 GeV (-0.5)<br />
0 10 20 30 40 50<br />
Nch 60 70 80 90 100
Tensi<strong>on</strong> between Tevatr<strong>on</strong> and LHC data?<br />
Rick Field: if p⊥0(ECM) ∝ Eǫ CM tuned to LHC data,<br />
then gives too much UE activity at Tevatr<strong>on</strong><br />
(⇒ need higher p⊥0 to compensate)<br />
<strong>PYTHIA</strong> Tune Z1<br />
P T0 (GeV/c)<br />
3.5<br />
3.0<br />
2.5<br />
2.0<br />
1.5<br />
1.0<br />
RDF Very Preliminary<br />
CMS 900 GeV<br />
MPI Cut-Off P T0(W cm)<br />
CDF 1.96 TeV<br />
Tune Z1<br />
CMS 7 TeV<br />
0 2000 4000 6000 8000 10000 12000<br />
Center-of-Mass Energy W cm (GeV)<br />
Pick some key LHC data sets, use Tune 2C as starting point:<br />
• slightly dampen diffractive cross secti<strong>on</strong> (ATLAS)<br />
• <strong>on</strong>ly vary MPI and colour rec<strong>on</strong>necti<strong>on</strong> parameters
New tunes<br />
. . . while waiting for more LHC data c<strong>on</strong>veniently implemented in Rivet.<br />
Parameter 2C 2M 4C 4Cx<br />
SigmaProcess:alphaSvalue 0.135 0.1265 0.135 0.135<br />
SpaceShower:rapidityOrder <strong>on</strong> <strong>on</strong> <strong>on</strong> <strong>on</strong><br />
SpaceShower:alphaSvalue 0.137 0.130 0.137 0.137<br />
SpaceShower:pT0Ref 2.0 2.0 2.0 2.0<br />
MultipleInteracti<strong>on</strong>s:alphaSvalue 0.135 0.127 0.135 0.135<br />
MultipleInteracti<strong>on</strong>s:pT0Ref 2.320 2.455 2.085 2.15<br />
MultipleInteracti<strong>on</strong>s:ecmPow 0.21 0.26 0.19 0.19<br />
MultipleInteracti<strong>on</strong>s:bProfile 3 3 3 4<br />
MultipleInteracti<strong>on</strong>s:expPow 1.60 1.15 2.00 N/A<br />
MultipleInteracti<strong>on</strong>s:a1 N/A N/A N/A 0.15<br />
BeamRemnants:rec<strong>on</strong>nectRange 3.0 3.0 1.5 1.5<br />
SigmaDiffractive:dampen off off <strong>on</strong> <strong>on</strong><br />
SigmaDiffractive:maxXB N/A N/A 65 65<br />
SigmaDiffractive:maxAX N/A N/A 65 65<br />
SigmaDiffractive:maxXX N/A N/A 65 65<br />
R. Corke & TS, JHEP 03 (2011) 032, JHEP 05 (2011) 009
Tune 4C now describes LHC data reas<strong>on</strong>ably well:<br />
1/N ev dN ev / dN ch<br />
<br />
10 2<br />
10 1<br />
10 0<br />
10 -1<br />
10 -2<br />
10 -3<br />
10 -4<br />
10 -5<br />
10 -6<br />
10 -7<br />
1.4<br />
1.2<br />
1.0<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
0.0<br />
ATLAS 7.00 TeV (x 100)<br />
Pythia 7.00 TeV (x 100)<br />
ATLAS 900 GeV<br />
Pythia 900 GeV<br />
0 10 20 30 40 50<br />
Nch 60 70 80 90 100<br />
ATLAS 7.00 TeV<br />
Pythia 7.00 TeV<br />
ATLAS 900 GeV<br />
Pythia 900 GeV<br />
0 2 4 6 8 10 12 14 16 18 20<br />
p T lead [GeV]<br />
Transverse<br />
regi<strong>on</strong><br />
1/N ev dN ev / dN ch<br />
[GeV]<br />
10 3<br />
10 2<br />
10 1<br />
10 0<br />
10 -1<br />
10 -2<br />
10 -3<br />
10 -4<br />
10 -5<br />
1.8<br />
1.6<br />
1.4<br />
1.2<br />
1<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
0<br />
ALICE 7.00 TeV (x 100)<br />
Pythia 7.00 TeV (x 100)<br />
ALICE 2.36 TeV (x 10)<br />
Pythia 2.36 TeV (x 10)<br />
ALICE 900 GeV<br />
Pythia 900 GeV<br />
0 10 20 30 40<br />
Nch 50 60 70 80<br />
ATLAS 7.00 TeV<br />
Pythia 7.00 TeV<br />
ATLAS 900 GeV (-0.5)<br />
Pythia 900 GeV (-0.5)<br />
0 10 20 30 40 50<br />
Nch 60 70 80 90 100
. . . but at the expense of Tevatr<strong>on</strong> agreement:<br />
dσ/dn ch<br />
MC/data<br />
1<br />
10 −1<br />
10 −2<br />
10 −3<br />
10 −4<br />
10 −5<br />
1.4<br />
1.2<br />
1<br />
0.8<br />
0.6<br />
Charged multiplicity at √ s = 1800 GeV, |η| < 1, pT > 0.4 GeV<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
CDF data<br />
<strong>PYTHIA</strong> 8.142 Tune 4C<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
0 5 10 15 20 25 30 35 40<br />
Future:<br />
• better understanding of data?<br />
• official/validated inclusi<strong>on</strong> in Rivet?<br />
• combined tune Tevatr<strong>on</strong> + LHC?<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
N ch<br />
〈N ch〉/dη dφ<br />
MC/data<br />
1<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
0<br />
1.4<br />
1.2<br />
1<br />
0.8<br />
0.6<br />
Transverse regi<strong>on</strong> charged particle density<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
CDF data<br />
<strong>PYTHIA</strong> 8.142 Tune 4C<br />
0 50 100 150 200 250 300 350 400<br />
pT(leading jet) / GeV
Further challenges from hadr<strong>on</strong> collider data<br />
Bose-Einstein r(N ch) ∝ N 1/3<br />
ch<br />
cannot be accommodated<br />
in <strong>PYTHIA</strong> effective descripti<strong>on</strong><br />
that worked at LEP<br />
R (fm) R (fm) R (fm)<br />
G<br />
out<br />
G<br />
side<br />
G<br />
l<strong>on</strong>g<br />
5<br />
1<br />
0<br />
5<br />
1<br />
0<br />
5<br />
1<br />
0<br />
PHENIX AuAu @ 200 AGeV<br />
STAR AuAu @ 200 AGeV<br />
STAR CuCu @ 200 AGeV<br />
2 4 6 8<br />
ALICE pp @ 7 TeV<br />
ALICE pp @ 0.9 TeV<br />
STAR pp @ 200 GeV<br />
2 4 6 8<br />
STAR AuAu @ 62 AGeV<br />
STAR CuCu @ 62 AGeV<br />
CERES PbAu @ 17.2 AGeV<br />
2 4 6 8<br />
2 4 6 8<br />
1/3<br />
〈 dN /dη<br />
〉<br />
Multiple overlapping fragmenting strings ⇒ dense hadr<strong>on</strong> gas!<br />
ch<br />
a)<br />
b)<br />
c)
Need more p ⊥ for K, p, Λ, . . . , relative to π ± :<br />
[GeV]<br />
〉<br />
T<br />
p<br />
〈<br />
1.2<br />
1<br />
0.8<br />
0.6<br />
0.4<br />
1.4<br />
1.2<br />
1<br />
0.8<br />
0.6<br />
200 GeV pp Minimum Bias<br />
Average Transverse Momentum vs Particle Mass (star-1)<br />
STAR<br />
Herwig++<br />
Perugia 2010<br />
Pythia 8<br />
STAR_2006_S6860818<br />
Herwig++ 2.4.2, Pythia 6.424, Pythia 8.145<br />
0.5 1 1.5<br />
mass [GeV]<br />
Ratio to STAR<br />
0.5 1 1.5<br />
mcplots.cern.ch<br />
MC / Data<br />
0<br />
S<br />
K<br />
MC / Data<br />
Λ<br />
MC / Data<br />
−<br />
Ξ<br />
1<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
1<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
1<br />
0.8<br />
0.6<br />
0.4<br />
0 1 2 3 4 5 6<br />
0 1 2 3 4 5 6<br />
7 TeV <strong>PYTHIA</strong>6 D6T<br />
7 TeV <strong>PYTHIA</strong>6 P0<br />
7 TeV <strong>PYTHIA</strong>8<br />
CMS<br />
0.9 TeV <strong>PYTHIA</strong>6 D6T<br />
0.9 TeV <strong>PYTHIA</strong>6 P0<br />
0.9 TeV <strong>PYTHIA</strong>8<br />
0.2<br />
0 1 2 3 4 5 6<br />
[GeV/c]<br />
p<br />
T
(K + + K - ) / (π + + π - )<br />
Latest π/K/p p ⊥ spectra from LHC<br />
• ALICE 900 GeV - inelastic events, |y| < 0.5, arXiv:1101.4110v3<br />
• ALICE 7 TeV - read from public talks (assume same acceptance as 900 GeV)<br />
• Compare against <strong>PYTHIA</strong> 8<br />
0.7<br />
0.6<br />
0.5<br />
0.4<br />
0.3<br />
0.2<br />
0.1<br />
0<br />
ALICE 900 GeV<br />
ALICE 7 TeV<br />
<strong>PYTHIA</strong> 900 GeV<br />
<strong>PYTHIA</strong> 7 TeV<br />
0 0.5 1 1.5 2 2.5<br />
p⊥ [GeV]<br />
Pilot study (R. Corke):<br />
model for rescattering as afterburner.<br />
No need to change particle compositi<strong>on</strong> (significantly),<br />
<strong>on</strong>ly (partial) collective flow?<br />
(p + + p - ) / (π + + π - )<br />
0.7<br />
0.6<br />
0.5<br />
0.4<br />
0.3<br />
0.2<br />
0.1<br />
0<br />
ALICE 900 GeV<br />
ALICE 7 TeV<br />
<strong>PYTHIA</strong> 900 GeV<br />
<strong>PYTHIA</strong> 7 TeV<br />
0 0.5 1 1.5 2 2.5<br />
p⊥ [GeV]
Final-state hadr<strong>on</strong> scattering<br />
Final-state hadr<strong>on</strong> scattering<br />
• Low-energy scatterings in CM frame and boost back<br />
• Higher masses take bigger kick, c.f. collective flow in heavy i<strong>on</strong><br />
• Ideally assign producti<strong>on</strong> vertices to outgoing hadr<strong>on</strong>s and follow path<br />
Simple model based <strong>on</strong> distance in y − φ space<br />
• High-p⊥ hadr<strong>on</strong>s in jets formed at later times and less likely to scatter<br />
• Scattering probability<br />
Pij = k σel<br />
ij (s)<br />
σ el<br />
max<br />
⎛<br />
max ⎝0, 1 − ∆R2 ij<br />
⎞<br />
R2 ⎠<br />
max<br />
• Order scatterings based <strong>on</strong> “relative transverse velocity”<br />
v ⊥ij =<br />
<br />
<br />
p⊥i <br />
m⊥i<br />
− p ⊥j<br />
m ⊥j<br />
• Dominated by pi<strong>on</strong>s, so focus <strong>on</strong> ππ, πK and πp (σ el<br />
max ∼ 200 mb)<br />
• Most scatterings close to threshold
σ el [mb]<br />
dσ el / d cos(θ)<br />
120<br />
100<br />
80<br />
60<br />
40<br />
20<br />
120<br />
100<br />
Elastic scattering cross secti<strong>on</strong>s<br />
Isospin partial-wave parameterisati<strong>on</strong>s (no Columb correcti<strong>on</strong>s)<br />
π + π -<br />
π + π 0<br />
π + π +<br />
0<br />
0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2<br />
ECM [GeV]<br />
80<br />
60<br />
40<br />
20<br />
1.0 * 0.40 GeV<br />
0.5 * 0.77 GeV<br />
1.0 * 1.10 GeV<br />
0<br />
-1 -0.8-0.6-0.4-0.2 0 0.2 0.4 0.6 0.8 1<br />
cos(θ)<br />
σ el [mb]<br />
dσ el / d cos(θ)<br />
80<br />
70<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
0<br />
80<br />
70<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
π + K -<br />
π 0 K +<br />
π + K +<br />
0.6 0.8 1 1.2 1.4 1.6 1.8 2<br />
ECM [GeV]<br />
1.0 * 0.80 GeV<br />
0.5 * 0.89 GeV<br />
1.0 * 1.10 GeV<br />
0<br />
-1 -0.8-0.6-0.4-0.2 0 0.2 0.4 0.6 0.8 1<br />
cos(θ)<br />
σ el [mb]<br />
dσ el / d cos(θ)<br />
250<br />
200<br />
150<br />
100<br />
70<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
50<br />
0<br />
π + p<br />
π 0 p<br />
π - p<br />
1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6<br />
ECM [GeV]<br />
1.0 * 1.10 GeV<br />
0.3 * 1.23 GeV<br />
1.0 * 1.40 GeV<br />
0<br />
-1 -0.8-0.6-0.4-0.2 0 0.2 0.4 0.6 0.8 1<br />
cos(θ)
Preliminary results<br />
Results from simple model (7 TeV)<br />
• Perform scattering after first round of hadr<strong>on</strong> decays (η → π)<br />
• HS1 - j = 0.1, Rmax = 1.0, flat cross secti<strong>on</strong> + isotropic scatterings<br />
• HS2 - j = 1.0, Rmax = 1.0, parameterised cross secti<strong>on</strong>s<br />
(K + + K - ) / (π + + π - )<br />
0.5<br />
0.4<br />
0.3<br />
0.2<br />
0.1<br />
0<br />
ALICE<br />
<strong>PYTHIA</strong> default<br />
<strong>PYTHIA</strong> HS1<br />
<strong>PYTHIA</strong> HS2<br />
0 0.5 1 1.5 2 2.5<br />
(p + + p - ) / (π + + π - )<br />
0.5<br />
0.4<br />
0.3<br />
0.2<br />
0.1<br />
0<br />
ALICE<br />
<strong>PYTHIA</strong> default<br />
<strong>PYTHIA</strong> HS1<br />
<strong>PYTHIA</strong> HS2<br />
0 0.5 1 1.5 2 2.5<br />
Cross secti<strong>on</strong>s are small, limiting effect<br />
• Less multiplicity at 900 GeV → less effect<br />
• Increase in Rmax tends to give pairs <strong>with</strong> higher invariant mass,<br />
but cross secti<strong>on</strong>s still low in this regi<strong>on</strong><br />
Preliminary results not so promising. . .
. . . but, whatever the correct explanati<strong>on</strong>,<br />
the road leads to more complexity, not less:<br />
Another example: matching to NLO, NNLO<br />
The need for experiment ↔ generators ↔ theory remains.<br />
PDF<br />
ME<br />
MPI<br />
ISR<br />
FSR<br />
BR<br />
CR<br />
Hadr.<br />
Decays<br />
Rescattering<br />
BE<br />
Unknown?
Summary and Outlook<br />
• <strong>PYTHIA</strong> 6 is winding down:<br />
⋆ is supported but not developed;<br />
⋆ still main opti<strong>on</strong> for current run (sigh),<br />
⋆ but not after l<strong>on</strong>g shutdown 2013!<br />
• <strong>PYTHIA</strong> 8 is the natural successor,<br />
⋆ is (sadly!) not yet quite up to speed in all respects,<br />
⋆ but for much already better than <strong>PYTHIA</strong>6 ,<br />
⋆ is starting to have competitive tunes,<br />
⋆ and will c<strong>on</strong>tinue to move ahead.<br />
⋆ Currently versi<strong>on</strong> 8.150 <strong>with</strong> the features menti<strong>on</strong>ed here.<br />
• Advise to experimentalists:<br />
⋆ step up <strong>PYTHIA</strong> 8 usage to gain experience;<br />
⋆ if you want new features then be prepared to use <strong>PYTHIA</strong> 8;<br />
⋆ provide feedback, both what works and what does not;<br />
⋆ do your own tunes to data and tell outcome.<br />
News list: http://www.hepforge.org/lists/listinfo/pythia8-announce<br />
The work is never d<strong>on</strong>e!