Event Generator Physics
Part 2: Hard processes and parton showers
Torbj¨ orn Sj¨ ostrand
Theoretical Particle Physics
Department of Astronomy and Theoretical Physics Lund University
S¨olvegatan 14A, SE-223 62 Lund, Sweden
DK–PI Summer School 2022, Neusiedl, Austria
Multijets – the need for Higher Orders
2→ 6 process or 2 → 2 dressed up by bremsstrahlung!?
Perturbative QCD
Perturbative calculations⇒ Matrix Elements.
Improved calculational techniques allows
? more legs (= final-state partons)
? more loops (= virtual partons not visible in final state) but with limitations, especially for loops.
Parton Showers:
approximations to matrix element behaviour,
most relevant for multiple emissions at low energies and/or angles.
Main topic of this lecture.
Matching and Merging:
methods to combine matrix elements (at high scales) with parton showers (at low scales),
with a consistent and smooth transition.
Huge field at LHC.
The basic QCD processes
Six basic 2→ 2 QCD processes:
qq0 → qq0 qq→ q0q0 qq→ gg qg→ qg gg→ qq gg→ gg Mandelstam variables
ˆ
s = (p1+ p2)2 = (p3+ p4)2 tˆ = (p1− p3)2 = (p2− p4)2 ˆ
u = (p1− p4)2 = (p2− p3)2
In rest frame, massless limit: m1 = m2 = m3 = m4 = 0 ˆ
s = ECM2 ˆt =−sˆ
2(1− cos ˆθ) ˆ
u =−sˆ
2(1 + cos ˆθ) ˆ
s + ˆt + ˆu = 0
Closeup: qg → qg
Considerq(1) g(2)→ q(3) g(4):
t : pg∗ = p1− p3 ⇒ m2g∗ = (p1− p3)2 = ˆt ⇒ dˆσ/dˆt ∼ 1/ˆt2 u : pq∗ = p1− p4 ⇒ m2q∗ = (p1− p4)2= ˆu⇒ dˆσ/dˆt ∼ −1/ˆsˆu s : pq∗ = p1+ p2 ⇒ mq2∗ = (p1+ p2)2 = ˆs ⇒ dˆσ/dˆt ∼ 1/ˆs2 Contribution of each sub-graph is gauge-dependent,
only sum is well-defined:
dˆσ
dˆt = πα2s ˆ s2
sˆ2+ ˆu2 tˆ2 +4
9 ˆ s (−ˆu)+4
9 (−ˆu)
ˆ s
Composite beams
In reality all beams are composite:
p : q, g, q, . . . e−:e−, γ, e+, . . . γ : e±, q, q, g
Factorization σAB =X
i,j
ZZZ
fi(A)(x1, Q2) fj(B)(x2, Q2)dˆσij
dˆt dx1dx2dˆt x : momentum fraction, e.g. pi = x1pA; pj = x2pB
Q2: factorization scale, “typical momentum transfer scale”
Factorization only proven for a few cases, likeγ∗/Z0 prodution, and strictly speaking not correct e.g. for jet production,
but good first approximation and unsurpassed physics insight.
In the beginning: Electrodynamics
An electrical charge, say an electron, is surrounded by a field:
For a rapidly moving charge
this field can be expressed in terms of an equivalent flux of photons:
dnγ ≈ 2αem π
dθ θ
dω ω Equivalent Photon Approximation, or method of virtual quanta (e.g. Jackson) (Bohr; Fermi; Weisz¨acker, Williams∼1934)
e−
e−
e−
e−
.
θ: collinear divergence, saved by me> 0 in full expression. ω: true divergence, nγ∝R
dω/ω =∞, but Eγ ∝R
ω dω/ω finite.
These are virtual photons: continuously emitted and reabsorbed.
In the beginning: Electrodynamics
An electrical charge, say an electron, is surrounded by a field:
For a rapidly moving charge
this field can be expressed in terms of an equivalent flux of photons:
dnγ ≈ 2αem π
dθ θ
dω ω Equivalent Photon Approximation, or method of virtual quanta (e.g. Jackson) (Bohr; Fermi; Weisz¨acker, Williams∼1934)
e−
e− e−
e−
.
θ: collinear divergence, saved by me> 0 in full expression. ω: true divergence, nγ∝R
dω/ω =∞, but Eγ ∝R
ω dω/ω finite. These are virtual photons: continuously emitted and reabsorbed.
In the beginning: Electrodynamics
An electrical charge, say an electron, is surrounded by a field:
For a rapidly moving charge
this field can be expressed in terms of an equivalent flux of photons:
dnγ ≈ 2αem π
dθ θ
dω ω Equivalent Photon Approximation, or method of virtual quanta (e.g. Jackson) (Bohr; Fermi; Weisz¨acker, Williams∼1934)
e−
e− e−
e−
.
θ: collinear divergence, saved by me> 0 in full expression.
ω: true divergence, nγ ∝R
dω/ω =∞, but Eγ ∝R
ω dω/ω finite.
These are virtual photons: continuously emitted and reabsorbed.
In the beginning: Bremsstrahlung
When an electron is kicked into a new direction, the field does not have time fully to react:
e−
Initial State Radiation (ISR):
part of it continues ∼ in original direction of e Final State Radiation (FSR):
the field needs to be regenerated around outgoing e, and transients are emitted ∼ around outgoing e direction Emission rate provided by equivalent photon flux in both cases.
Approximate cutoffs related to timescale of process:
the more violent the hard collision, the more radiation!
In the beginning: Exponentiation
AssumeP
Eγ Ee such that energy-momentum conservation is not an issue. Then
dPγ =dnγ ≈ 2αem π
dθ θ
dω ω is the probability to find a photon atω and θ, irrespectively of which other photons are present.
Uncorrelated⇒ Poissonian number distribution:
Pi = hnγii i ! e−hnγi with
hnγi = Z θmax
θmin
Z ωmax
ωmin
dnγ ≈ 2αem π ln
θmax θmin
ln
ωmax ωmin
Note thatR
dPγ =R
dnγ> 1 is not a problem:
proper interpretation is that many photons are emitted.
Exponentiation: reinterpretation of dPγ into Poissonian.
So how is QCD the same?
A quark is surrounded by a gluon field dPg =dng ≈ 8αs
3π dθ
θ dω
ω i.e. only differ by substitution αem→ 4αs/3.
An accelerated quark emits gluons with collinear and soft divergences, and as Initial andFinal State Radiation.
e−
q
Typically hngi =R
dng 1 since αs αem
⇒ even more pressing need for exponentiation.
So how is QCD different?
QCD is non-Abelian, so a gluon is charged and is surrounded by its own field:
emission rate 4αs/3→ 3αs, field structure more complicated, interference effects more important.
αs(Q2) diverges for Q2 → Λ2QCD, with ΛQCD ∼ 0.2 GeV = 1 fm−1. Confinement: gluons below ΛQCD not resolved ⇒ de facto cutoffs.
.
Unclear separation between
“accelerated charge” and “emitted radiation”:
many possible Feynman graphs≈ histories.
The Parton-Shower Approach
2→ n = (2 → 2) ⊕ ISR ⊕ FSR
FSR = Final-State Radiation = timelike shower Qi2∼ m2 > 0 decreasing
ISR = Initial-State Radiation = spacelike showers Qi2∼ −m2> 0 increasing
Why “time”like and “space”like?
Consider four-momentum conservation in a branching a→ b c p⊥a= 0 ⇒ p⊥c =−p⊥b
p+= E + pL ⇒ p+a= p+b+ p+c
p−= E− pL ⇒ p−a= p−b+ p−c Define p+b = z p+a, p+c = (1− z) p+a
Use p+p−= E2− pL2 = m2+ p⊥2 m2a+ p2⊥a
p+a
= m2b+ p⊥b2 z p+a
+ m2c+ p2⊥c (1− z) p+a
⇒ ma2= mb2+ p2⊥
z +m2c+ p2⊥ 1− z = m2b
z + m2c
1− z + p⊥2 z(1− z) Final-state shower: mb= mc = 0⇒ ma2= p
2
⊥
z(1−z) > 0⇒ timelike Initial-state shower: ma = mc = 0⇒ m2b=−1p−z2⊥ < 0⇒ spacelike
Doublecounting
Do not doublecount: 2→ 2 = most virtual = shortest distance (detailed handling of borders⇒ match & merge)
Final-state radiation
Standard process e+e−→ qqg by two Feynman diagrams:
xi = 2Ei
Ecm x1+x2+x3= 2
dσME
σ0
=
2παs 43 (1−xx12+x221)(1−x2)
dx
1dx
2Convenient (but arbitrary) subdivision to “split” radiation: 1
(1− x1)(1− x2)
(1− x1) + (1− x2)
x3 = 1
(1− x2)x3 + 1 (1− x1)x3
Final-state radiation
Standard process e+e−→ qqg by two Feynman diagrams:
xi = 2Ei
Ecm x1+x2+x3= 2
dσME
σ0
=
2παs 43 (1−xx12+x221)(1−x2)
dx
1dx
2Convenient (but arbitrary) subdivision to “split” radiation:
1
(1− x1)(1− x2)
(1− x1) + (1− x2)
x3 = 1
(1− x2)x3 + 1 (1− x1)x3
From matrix elements to parton showers
Rewrite for x2 → 1, i.e. q–g collinear limit:
1− x2 = m213 Ecm2 = Q2
Ecm2 ⇒ dx2= dQ2 Ecm2 define z as fraction q retains
in branchingq→ qg
x1 ≈ z ⇒ dx1 ≈ dz x3 ≈ 1 − z
⇒ dP =dσ σ0
= αs
2π dx2
(1− x2) 4 3
x22+ x12
(1− x1) dx1 ≈ αs
2π dQ2
Q2 4 3
1 + z2 1− z dz In limit x1 → 1 same result, but for q → qg.
dQ2/Q2 =dm2/m2: “mass (or collinear) singularity”
dz/(1− z) = dω/ω “soft singularity”
The DGLAP equations
Generalizes to
DGLAP (Dokshitzer–Gribov–Lipatov–Altarelli–Parisi)
dPa→bc = αs 2π
dQ2
Q2 Pa→bc(z)dz Pq→qg = 4
3 1 + z2
1− z
Pg→gg = 3(1− z(1 − z))2 z(1− z) Pg→qq = nf
2 (z2+ (1− z)2) (nf =no. of quark flavours) Universality: any matrix element reduces to DGLAP in collinear limit.
e.g. dσ(H0 → qqg)
dσ(H0 → qq) = dσ(Z0→ qqg)
dσ(Z0→ qq) in collinear limit
The iterative structure
Generalizes to many consecutive emissions if strongly ordered, Q12 Q22 Q32. . . (≈ time-ordered).
To cover “all” of phase space use DGLAP in whole region Q12> Q22 > Q32. . ..
Iteration gives final-state parton showers:
Need soft/collinear cuts to stay away from nonperturbative physics.
Details model-dependent, but around 1 GeV scale.
The Sudakov form factor – 1
Time evolution, conservation of total probability:
P(no emission) = 1 − P(emission).
Multiplicativeness, with Ti = (i/n)T , 0≤ i ≤ n:
Pno(0≤ t < T ) = lim
n→∞
nY−1 i =0
Pno(Ti ≤ t < Ti +1)
= lim
n→∞
nY−1 i =0
(1− Pem(Ti ≤ t < Ti +1))
= exp − limn
→∞
n−1
X
i =0
Pem(Ti ≤ t < Ti +1)
!
= exp
− Z T
0
dPem(t) dt dt
=⇒ dPfirst(T ) = dPem(T )exp
− Z T
0
dPem(t) dt dt
cf. radioactive decay in lecture 1.
The Sudakov form factor – 2
Expanded, with Q ∼ 1/t (Heisenberg) dPa→bc = αs
2π dQ2
Q2 Pa→bc(z)dz
× exp
−X
b,c
Z Qmax2 Q2
dQ02 Q02
Z αs
2πPa→bc(z0)dz0
where the exponent is (one definition of) the Sudakov form factor A given parton can only branch once, i.e. if it did not already do so Note thatP
b,c
R R dPa→bc ≡ 1 ⇒ convenient for Monte Carlo (≡ 1 if extended over whole phase space, else possibly nothing happens before you reach Q0 ≈ 1 GeV).
The Sudakov form factor – 3
Sudakov regulates singularity for first emission . . .
. . . but in limit of repeated soft emissionsq→ qg (but no g → gg) one obtains the same inclusive Q emission spectrum as for ME, i.e. divergent ME spectrum
⇐⇒ infinite number of PS emissions
More complicated in reality:
energy-momentum conservation effects big since αs big, so hard emissions frequent
g→ gg branchings leads to accelerated multiplication of partons
The ordering variable
In the evolution with
dPa→bc = αs 2π
dQ2
Q2 Pa→bc(z)dz Q2 orders the emissions (memory).
If Q2= m2 is one possible evolution variable then Q02= f (z)Q2 is also allowed, since
d(Q02, z) d(Q2, z) =
∂Q02
∂Q2
∂Q02
∂z ∂z
∂Q2 ∂z
∂z
=
f (z) f0(z)Q2
0 1
= f (z)
⇒ dPa→bc = αs 2π
f (z)dQ2
f (z)Q2 Pa→bc(z)dz = αs 2π
dQ02
Q02 Pa→bc(z)dz Q02= Ea2θa2→bc ≈ m2/(z(1− z)); angular-ordered shower Q02= p⊥2 ≈ m2z(1− z); transverse-momentum-ordered
Coherence
QED: Chudakov effect (mid-fifties)
QCD: colour coherence for soft gluon emission
solved by • requiringemission angles to be decreasing or • requiringtransverse momenta to be decreasing
Coherence
QED: Chudakov effect (mid-fifties)
QCD: colour coherence for soft gluon emission
solved by • requiringemission angles to be decreasing or • requiringtransverse momenta to be decreasing
The dipole picture
1→ 2 branching = replace m = 0 parton by pair with m > 0.
Breaks energy–momentum conservation.
Herwig angular-ordered shower: post-facto rescaling machinery.
Alternative: dipole picture (most common, 3 variants in PYTHIA).
2→ 3 parton branching, or 1 → 2 colour dipole branching.
Can be viewed as radiator a→ bc with recoiler r.
Quark vs. gluon jets
Pg→gg
Pq→qg ≈ Nc
CF = 3 4/3 = 9
4 ≈ 2
⇒ gluon jets are softer and broader than quark ones (also helped by hadronization models, lecture 4).
(GeV/c) Jet PT
50 100 200 300 1000
〉chN〈
5 10 15 20 25
= 7 TeV s pp
Data |y| < 1 Data 1 < |y| < 2 Gluon Jets (Pythia Tune Z2) Quark Jets (Pythia Tune Z2)
L dt = 36 pb-1
∫
CMS
(GeV/c) Jet PT
50 100 200 300 1000
〉2Rδ〈
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035
= 7 TeV s pp
Data |y| < 1 Data 1 < |y| < 2 Gluon Jets (Pythia Tune Z2) Quark Jets (Pythia Tune Z2)
L dt = 36 pb-1
∫
CMS
Note transition g jets→ q jets for increasing p⊥.
Parton Distribution Functions
Hadrons are composite, with time-dependent structure:
fi(x, Q2) = number density of partons i at momentum fraction x and probing scale Q2.
Linguistics (example):
F2(x, Q2) = X
i
ei2xfi(x, Q2) structure function parton distributions
PDF evolution
Initial conditions at small Q02 unknown: nonperturbative.
Resolution dependence perturbative, by DGLAP:
DGLAP (Dokshitzer–Gribov–Lipatov–Altarelli–Parisi) dfb(x, Q2)
d(ln Q2) =X
a
Z 1 x
dz
z fa(y, Q2) αs 2πPa→bc
z = x
y
DGLAP already introduced for (final-state) showers:
dPa→bc = αs 2π
dQ2
Q2 Pa→bc(z)dz Same equation, but different context:
dPa→bc is probability for the individual parton to branch; while dfb(x, Q2) describes how the ensemble of partons evolve by the branchings of individual partons as above.
Initial-State Shower Basics
• Parton cascades in p are continuously born and recombined.
• Structure at Q is resolved at a time t ∼ 1/Q before collision.
• A hard scattering at Q2 probes fluctuations up to that scale.
• A hard scattering inhibits full recombination of the cascade.
• Convenient reinterpretation:
Forwards vs. backwards evolution
Event generation could be addressed by forwards evolution:
pick a complete partonic set at low Q0 and evolve, consider collisions at different Q2 and pick byσ of those.
Inefficient:
1 have to evolve and check for all potential collisions, but 99.9. . . % inert
2 impossible (or at least very complicated) to steer the production, e.g. of a narrow resonance (Higgs)
Backwards evolution is viable and∼equivalent alternative:
start at hard interaction and trace what happened “before”
Backwards evolution master formula
Monte Carlo approach, based on conditional probability : recast dfb(x, Q2)
dt =X
a
Z 1
x
dz
z fa(x0, Q2) αs
2π Pa→bc(z) with t = ln(Q2/Λ2) andz = x/x0 to
dPb= dfb
fb =|dt| X
a
Z
dz x0fa(x0, t) xfb(x, t)
αs
2π Pa→bc(z) then solve for decreasing t, i.e. backwards in time,
starting at high Q2 and moving towards lower, with Sudakov form factor exp(−R
dPb).
Extra factor x0fa/xfb relative to final-state equations.
Initial- vs. final-state showers
Both controlled by same evolution equations dPa→bc = αs
2π dQ2
Q2 Pa→bc(z)dz · (Sudakov) but
Final-state showers:
Q2 timelike (∼ m2)
decreasing E, m2, θ both daughters m2 ≥ 0 physics relatively simple
⇒ “minor” variations:
Q2, shower vs. dipole, . . .
Initial-state showers:
Q2 spacelike (≈ −m2)
decreasing E , increasing Q2, θ one daughter m2 ≥ 0, one m2 < 0 physics more complicated
⇒ more formalisms:
DGLAP, BFKL, CCFM, GLR, . . .
Combining FSR with ISR
Separate processing of ISR and FSR misses interference (∼ colour dipoles)
ISR+FSR add coherently in regions of colour flow and destructively else in “normal” shower by azimuthal anisotropies automatic in dipole (by proper boosts)
Combining FSR with ISR
Separate processing of ISR and FSR misses interference (∼ colour dipoles)
ISR+FSR add coherently in regions of colour flow and destructively else in “normal” shower by azimuthal anisotropies automatic in dipole (by proper boosts)
Next-to-leading log showers
dPg=dng ≈ 8αs 3π
dθ θ
dω
ω 7→ αsL2 gives leading-log answer Pn ∝ (αsL2)n=αnsL2n.
Resummation/exponentiation gives Sudakov P0 ∝ exp(−αsL2).
(Transverse momentum cuts bothθ and ω ⇒ αnsLn.)
More careful handling of kinematics,αs running, splitting kernels (alsog→ ggg), etc., give subleading corrections ∝αnsL2n−1. All showers have some elements of NLL, e.g. momentum conservation, but some dedicated ongoing projects:
PanScales (Salam et al.) Deductor (Nagy, Soper) Herwig 7 (Pl¨atzer et al.) Vincia (Skands et al.) Dire (H¨oche, Prestel)
PanScales
PanScales is a dipole shower, but with several special choices.
Set of possible evolution scales, withβ = 0 like p⊥. Recoil local inside dipole or global in whole system.
Careful splitting of emission on two sides of dipole.
Higher-order splitting kernels and αs.
Some more. 4
FIG. 2. Left: ratio of the cumulative y23distribution from several showers divided by the NLL answer, as a function of
↵sln y23/2, for ↵s! 0. Right: summary of deviations from NLL for many shower/observable combinations (either ⌃shower(↵s! 0, ↵sL = 0.5)/⌃NLL 1 or (Nshowersubjet(↵s! 0, ↵sL2= 5)/NNLLsubjet 1)/p↵s). Red squares indicate clear NLL failure; amber triangles indicate NLL fixed-order failure that is masked at all orders; green circles indicate that all NLL tests passed.
Fig. 1.
The left-hand plot of Fig. 1 shows the Pythia8 dipole algorithm (not designed as NLL accurate), while the middle plot shows our PanGlobal shower with = 0.
The dipole result is clearly not independent of 12
for ↵s! 0, with over 60% discrepancies, extending the fixed-order conclusions of Ref. [37]. The discrepancy is only' 30% for gg events (not shown in Fig. 1), and the di↵erence would, e.g., skew machine learning [67] for quark/gluon discrimination. PanGlobal is independent of 12. The right-hand plot shows the ↵s! 0 limit for multiple showers. The overall pattern is as expected:
PanLocal works for = 0.5, but not = 0, demon- strating that with ktordering it is not sufficient just to change the dipole partition to get NLL accuracy. Pan- Global works for = 0 and = 0.5. (Showers that coincide for ↵s! 0, e.g. Dire v1 and Pythia8, typically di↵er at finite ↵s, reflecting NNLL di↵erences.)
Next, we consider a range of more standard observ- ables at NLL accuracy. They include the Cambridge py23resolution scale [68]; two jet broadenings, BTand BW [69]; fractional moments, FC1 obs, of the energy- energy correlations [47]; the thrust [70, 71], and the max- imum ui= kti/Qe obs|⌘i|among primary Lund declus- terings i. Each of these is sensitive to soft-collinear ra- diation as kt/Qe obs|⌘|, with the obs values shown in Fig. 2 (right). Additionally, the scalar sum of the trans- verse momenta in a rapidity slice [72], of full-width 2, is useful to test non-global logarithms (NGLs). These ob- servables all have the property that their distribution at NLL can be written as [47, 53, 72–74]
⌃(↵s, ↵sL) = exp⇥
↵s1g1(↵sL) + g2(↵sL) +O ↵nsLn 1⇤ , (6) where ⌃ is the fraction of events where the observable is smaller than eL (g1 = 0 for the rapidity slice kt).
We also consider the k-algorithm [75] subjet multiplic-
ity [76], [51]§ 5.
Fig. 2 (left) illustrates our all-order tests of the shower for one observable, py23. It shows the ratio of the ⌃ as calculated with the shower to the NLL result, as a function of ↵sln py23in the limit of ↵s! 0. The stan- dard dipole algorithms disagree with the NLL result, by up to 20%. This is non-negligible, though smaller than the disagreement in Fig. 1, because of the azimuthally averaged nature of the py23observable. In contrast the PanGlobal and PanLocal( = 0.5) showers agree with the NLL result to within statistical uncertainties.
Fig. 2 (right) shows an overall summary of our tests. The position of each point shows the result of
⌃shower(↵s! 0, ↵sL = 0.5)/⌃NLL 1 or (Nshowersubjet(↵s! 0, ↵sL2= 5)/NNLLsubjet 1)/p↵s. If it di↵ers from 0, the point is shown as a red square. In some cases (amber tri- angles) it agrees with 0, though an additional fixed-order analysis in a fixed-coupling toy shower [37] [51]§ 2 re- veals issues a↵ecting NLL accuracy, all involving hitherto undiscovered spurious super-leading logarithmic terms.1 Green circles in Fig. 2 (right) indicate that the shower/observable combination passes all of our NLL tests, both at all orders and in fixed-order expansions.
The four shower algorithms designed to be NLL accurate pass all the tests. These are the PanLocal shower (dipole and antenna variants) with = 12 and the PanGlobal shower with = 0 and =12.
1Such terms, (↵sL)n(↵sL2)pin ln ⌃, starting typically for n = 3 (sometimes 2), p 1, appear for traditional ktordered dipole showers for global (obs> 0) and non-global observables [51]§ 3.
Terms of this kind can generically exist [77–79], but not at leading-colour or for pure final-state processes with rIRC [47]
safe observables. In many cases, the spurious super-leading log- arithms appear to resum to mask any disagreement with NLL.
Correct NLL answer from analytical resummation variable by variable.
Torbj¨orn Sj¨ostrand Event Generator Physics 2 slide 34/44
Matrix elements vs. parton showers
ME : Matrix Elements
+ systematic expansion in αs (‘exact’) + powerful for multiparton Born level + flexible phase space cuts
− loop calculations very tough
− negative cross section in collinear regions
⇒ unpredictive jet/event structure
− no easy match to hadronization PS : Parton Showers
− approximate, to LL (or NLL)
− main topology not predetermined
⇒ inefficient for exclusive states + process-generic ⇒ simple multiparton + Sudakov form factors/resummation
⇒ sensible jet/event structure + easy to match to hadronization
Matrix elements and parton showers
Recall complementary strengths:
• ME’s good for well separated jets
• PS’s good for structure inside jets Marriage desirable! But how?
Problems: • gaps in coverage?
• doublecounting of radiation?
• Sudakov?
• NLO consistency?
Much work in last 20 – 30 years; impossible to cover it all.
Three main areas, in ascending order of complication:
1) Matrix-element correction with lowest-order nontrivial process 2) Merge leading-order multiparton process — vetoed showers 3) Match to next-to-leading order process —MC@NLO, POWHEG
Note unintuitive terminology match vs. merge⇒ M&M
Matrix Element Corrections (MEC)
= cover full phase space with smooth transition ME/PS.
Want to reproduce WME= 1 σ(LO)
dσ(LO + g) d(phasespace) by shower generation withWPS>WME + correction procedure
wanted
z }| { WME=
generated
z }| { WPS
correction
z }| { WME
WPS
• Exponentiate ME correction by shower Sudakov form factor:
WactualPS (Q2)=WME(Q2)exp − Z Qmax2
Q2
WME(Q02)dQ02
!
• Memory of shower remains in Q2 choice, i.e. “time” ordering.
• ME regularized: probability ≤ 1 instead of divergent.
• NLO correction simple for FSR, more messy for ISR:
replaceσ(LO)→ σ(NLO) in prefactor.
Vetoed Parton Showers (Merging)
Generic method to combine ME’s of several different orders to NLL accuracy; has become a standard tool for many studies Basic idea:
consider (differential) cross sectionsσ0, σ1, σ2, σ3, . . ., corresponding to a lowest-order process (e.g. W production), with more jets added to describe more complicated topologies, in each case to the respective leading order
σi, i ≥ 1, are divergent in soft/collinear limit
absent virtual corrections would have ensured “detailed balance”, i.e. an emission that adds toσi +1 subtracts from σi such virtual corrections correspond (approximately) to the Sudakov form factors of parton showers
so use shower routines to provide missing virtual corrections
⇒ rejection of events (especially) in soft/collinear regions
Simple merging example
Considere+e−→ qq, qqg, qqgg, with Qmax> Q > Qmin. Early matching without ME correction would read
Wqqg=WME(Q2)exp − Z Q2max
Q2
WPS(Q02)dQ02
!
For fixed-multiplicity topologies at scale Qmin: Wqq = exp −
Z Qmax2 Q2min
WqqPS
!
Wqqg = exp − Z Qmax2
Q2
WqqPS
!
WqqgME(Q2)exp − Z Q2
Qmin2
WqqgPS
!
Wqqgg = exp − Z Qmax2
Q21
WqqPS
!
exp − Z Q21
Q22
WqqgPS
!
WqqggME(Q12, Q22)
× exp − Z Q22
Q2min
WqqggPS
!
(with intermediate scales Q1> Q2) Regular shower allowed only below Q (unlike matching).
Veto scheme
1) Pick hard process, mixing according toσ0:σ1 :σ2 :. . ., above some ME cutoff(e.g. all p⊥i > p⊥0, all Rij > R0), with large fixedαs0
2) Reconstruct imagined shower history (in different ways) 3) Weight Wα =Q
branchings(αs(k⊥i2 )/αs0) ⇒ accept/reject CKKW-L:
4) Sudakov factor for non-emission on all lines above ME cutoff WSud =Q
“propagators00
Sudakov(k⊥beg2 , k⊥end2 ) 4a) CKKW : use NLL Sudakovs 4b) L: use trial showers
5) WSud ⇒ accept/reject 6) do shower, vetoing emissions above cutoff
MLM:
4) do parton showers 5) (cone-)cluster showered event
6) match partons and jets 7) if all partons are matched, and njet = nparton, keep the event, else discard it
Matching
Objectives:
Total rate should be accurate to NLO.
NLO results are obtained for all observables when (formally) expanded in powers of αs.
Hard emissions are treated as in the NLO computations.
Soft/collinear emissions are treated as in shower MC.
The matching between hard and soft emissions is smooth.
The outcome is a set of “normal” events.
Two commonly used schemes — MC@NLO and POWHEG — but intermediate or quite different ones could be possible.
Matching: MC@NLO
Madgraph5 aMC@NLO:
automated MC@NLO calculations for “arbitrary” process.
Matching: POWHEG
dσ = ¯B(v )dΦv
R(v, r ) B(v ) exp
− Z
p⊥
R(v, r0) B(v ) dΦ0r
dΦr
where
B(v ) = B(v ) + V (v ) +¯ Z
dΦr[R(v, r )− C (v, r)] . and
v, dΦv Born-level n-body variables and differential phase space r, dΦr extra n + 1-body variables and differential phase space B(v ) Born-level cross section
V (v ) Virtual corrections
R(v, r ) Real-emission cross section
C (v, r ) Conterterms for collinear factorization of parton densities.
Note thatR
[· · · dΦr]≡ 1 andR B(v )dΦ¯ v ≡ σNLO.
Summary
Overall production rate approximately set by lowest-order process.
Parton showers give multijets and subjet structure.
Sudakov factor is the crucial element for shower construction.
Recall “resummation” 1− x + O(x2)7→ exp(−x).
Dipole approach convenient for coherence and energy–momentum conservation.
ISR technically more complicated than FSR.
Matching&merging current frontline in the struggle to provide highest possible perturbative accuracy of event properties.