Perturbative QCD: from the

Tevatron to the LHC

R
Factorization scale
F
Renormalization scale
R
Short distance cross
section, calculated as
a perturbation series
in !
S
Physical cross
section
Parton distribution function
!
S
!
S
is small(ish) at high energies because of the
property of asymptotic freedom.
The role of LEP in determining the size of !
S
has
been crucial
G. Altarelli 1989 S. Kluth EPS, 2007
2006 World average !
s
(M
z
)

=0.11750.0011
S Kluth, hep-ex/0609020
!
s
known to ~1%!
!
s
from event shapes: Prehistory
In 1980 RKE, D.A. Ross and Terrano
considered jet shapes in e
+
e
-
annihilation.
We introduced the subtraction method for the
cancellation of real and virtual singularities.
We calculated the NLO corrections to the C-
parameter defined in terms of the eigenvalues
of the 3x3 matrix ". C is the coefficient of the
linear term in the characteristic equation.
C parameter
In 1980 the NLO corrections
were found to be large. The
dashed curve shows LO result
multiplied by (1+ 4.45 !
s
).
Dissertori et al, arXiv:0712.0327v2
NNLO results for thrust
A=LO, Farhi, 1977
B=NLO, Kunszt 1980
C=NNLO, Gehrmann-De Ridder, Gehrmann, Glover, Heinrich 2007
The thrust is defined as the maximum of
directed momentum
Defining #=1-T, perturbative expansion of thrust is
After many years of work the calculation of the O(!
s
3
) term was
completed by Gehrmann et al.
The first time a subtraction scheme has been implemented at NNLO.
Determination of !
s
from NNLO
Spread between different
observables is reduced because
of different event shape variables
have different NNLO
corrections.
Scale variation uncertainty
reduced by a factor 2 with
respect to NLO.
Scale variation error still the
largest.
Dissertori et al, arXiv 0712.0327
Resummation results for thrust
Expression for thrust contains large logarithms
for small #
Catani et al 1993
Becher, Schwartz, 2008
Becher, Schwartz, 2008
Matched result for !
s
in
N
3
LLA+NNLO
Becher and Schwartz, arXiv:0803.0342
! N
3
LLA resummation
performed using soft collinear
effective theory.
!Expansion of resummed
result near kinematic endpoint
provides a check of fixed
order NNLO calculation.
!For the first time the scale
variation error is not the
largest error.
Parton distribution functions
Measurement of the non-perturbative parton
distributions at lower energies allow extrapolations to
higher values of and lower values of x using the
DGLAP equation
The evolution kernel is calculable as a
perturbation series in !
s
LO NLO NNLO
NNLO is known completely. (Moch et al, hep-ph/0403192)
Comparison of H1 and Zeus
Some of the differences are understood (inclusion of BCDMS
at large x (ZEUS) ; inclusion of jet data for mid x gluon (H1))
arXiv:0903.3861
Projected parton model
uncertainties after HERAII
HERA for LHC workshop,hep-ph/0601012
and consequent improvement
on uncertainty of jet cross section
Why NLO?
Less sensitivity to unphysical
input scales, (eg.
renormalization and
factorization scales)
NLO first approximation in
QCD which gives an idea of
suitable choice for .
NLO has more physics, parton
merging to give structure in jets,
initial state radiation, more
species of incoming partons
enter at NLO.
A necessary prerequisite for
more sophisticated techniques
which match NLO with parton
showering.
In order to get ~10% accuracy
we need to include NLO.
Top total cross section at NLO
Short distance cross
section given by
Nason,Dawson,Ellis
Total cross section for top:
Analytic results (2008)
Czakon and
Mitov have
analytic results
for the total
cross section.
Results agree
with NDE fit
within stated
tolerance (1%).
Comparison with NDE fit for q-qbar
Comparison with NDE fit for g-g
Czakon-Mitov arXiv:0811.4119
Resummation of Threshold
Logarithms
NLO calculation can be used to determine coefficient in
resummed formula. Despite accuracy of fit , constant coefficient
C
3
was not well determined in fit.
Phenomenological impact on the cross section, not yet published, but it
will probably lead to a modest decrease ~1% (Nason, private communication).
Czakon-Mitov. arXiv 0811.4119
Resummation of Threshold Logs
Resummation performed in
moment space, separately
for each color channel
Top at LHC: Uncertainty budget
Best prediction (without the update in coefficient C
3
)
LO and NLO
Cacciari et al , arXiv 0804.2800
Top mass uncertainty --> 5 $m
t
/m
t
~ 5%
! Scale uncertainty dominant at LHC.
!Limited benefit for scale uncertainty from NLL resummation.
!Provides motivation for NNLO calculation, reduction of scale
uncertainty to 3%?, Moch-Uwer et al , arXiv 0804.1476
Progress on the
NNLO Top quark cross section
Motivation: Scale dependence is dominant error at LHC.
Standard candle for gg flux.
Loop-by-loop, Anastasiou, 0809.1355
Korner et al, arXiv:, 0802.0106, 0809.3980
2-loop amplitudes, qqbar
Czakon, arXiv:0803.1400
Tt+jet, Dittmaier arXiv:0810.0452
MCFM
A NLO parton level generator
pp ! W/Z
pp ! W+Z, WW, ZZ
pp ! W/Z + 1 jet
pp ! W/Z +2 jets
pp ! t W
pp ! tX (s&t channel)
pp ! tt
pp ! W/Z+H
pp (gg) ! H
pp!(gg) ! H + 1 jet
pp!(gg) ! H + 2 jets
pp(VV) ! H +2 jets
pp ! W/Z +b , W+c
pp ! W/Z +bb
.
Processes calculated at NLO, but no automatic procedure for including new processes.
Code available at http://mcfm.fnal.gov
Current version 5.4 (March 11, 2009)
The big picture
MCFM contains the best
predictions for many
processes of relevance for
Tevatron and LHC.
LHC will be a great
machine because of the
increase of both energy
and luminosity wrt to the
Tevatron.
Dramatic growth with
energy of gluon-induced
processes (eg tt).
W+n jet rates from CDF
MCFM RKE, Campbell
Both uncertainty on rates and deviation of Data/Theory from 1 are smaller than
other calculations. The ratio R agrees well for all theory calculations, but only
available from MCFM with small error for n%2.
D0 arXiv:0903.1748
" MCFM, LO and NLO agrees with data;
" shower-based generators show significant differences with data;
" matrix element + parton shower models agree in shape, but with
larger normalization uncertainties.
New Z + jets results from D0
Extension to multi-leg processes
At the LHC we are interested in processes with
many jets; these have standard model backgrounds
involving many legs.
The NLO calculation of multi-leg processes is
pressing because the dependence on the
unphysical scales is so strong.
We need both efficient methods to calculate tree
diagrams and efficient methods to calculate loops.
The calculation of one loop amplitudes
The classical paradigm for
the calculation of one-loop
diagrams was established
in 1979.
Complete calculation of
one-loop scalar integrals
Reduction of tensors one-
loop integrals to scalars.
Neither will be adequate for present-day purposes.
Basis set of scalar integrals
In the context of NLO calculations, scalar higher point functions, can always be
expressed as sums of box integrals. Passarino, Veltman - Melrose (65)
Any one-loop amplitude can be written as a linear sum of scalar
box-, triangle-, bubble- and tadpole-integrals.
Scalar hexagon can be written as a sum of six pentagons.
For the purposes of NLO calculations, the scalar pentagon can be written as a sum of
five boxes.
In addition to the finite integrals we need integrals containing infrared and collinear
divergences.
Scalar one-loop integrals
t Hooft and Veltmans integrals contain internal masses;
however in QCD many lines are (approximately) massless.
The consequent soft and collinear divergences are
regulated by dimensional regularization.
Analytic results are given for boxes, triangles, bubbles and
tadpoles, including the cases with one or more vanishing
internal masses at http://qcdloop.fnal.gov
Fortran 77 code is provided which calculates an arbitrary
scalar box, triangle, bubble or tadpole integral.
Problem of one-loop scalar integrals for NLO calculations
is completely solved numerically and analytically!
Basis set of sixteen divergent box integrals
RKE, Zanderighi
Dashed lines massless, lines of same
colour have same virtuality/mass
Example of box integral from
qcdloop.fnal.gov
Basis set of 16 basis integrals allows
the calculation of any divergent box
diagram.
Result given in the spacelike region.
Analytic continuation as usual by
s
ij
& s
ij
+ i '
Limit p
3
2
=0 can be obtained from this result, (limit p
2
2
=0 cannot)
Determination of coefficients of
scalar integrals
Feynman diagrams may not be the answer as the number of legs
increases. There are too many diagrams with cancellations
between them.
Semi-numerical methods based on unitarity offer great promise.
Semi-numerical because the integral containing the divergences
is determined analytically, but its coefficient is determined
numerically.
Unitarity for one-loop diagrams
Important steps include:-
First modern use of the idea Bern, Dixon,Kosower
Cuts w.r.t. to loop momenta give (box) coefficients
directly Cachazo, Britto, Feng
OPP tensor reduction scheme, Ossola, Pittau, Papadopoulos
Integrating the OPP procedure with unitarity Ellis, Giele,
Kunszt
D-dimensional unitarity Giele, Kunszt, Melnikov
Unitarity in D-dimensions
The theory contains divergences which we regulate
dimensionally. Divergences give poles as ' =(4-D)/2 ->0
Calculate the unitarity cuts numerically in integer
dimensions D>4. Internal degrees of freedom are taken to
be D
s
dimensional.
Dependence on D
s
is linear so we calculating in a two
different integer dimensions and extrapolate to '=0
Only the length of the loop momentum in the extra
dimension is relevant so we can treat the loop momentum as
five-dimensional.
Giele, Kunszt,Melnikov, arXiv 0801.2237
One loop calculation of pure
gluon amplitudes
Giele, Zanderighi arXiv:0805.2152
Time to calculate
one-loop amplitude
scales as N
9
as
expected. For small
numbers of legs
N=4,5,6 the times are
of the order of 10s
of milliseconds
4g:Ellis-Sexton(1985)
5g:Bern-Dixon-Kosower(1993)
6g:Ellis-Giele-Zanderighi(2006)
Generalized unitarity and
massive fermions
Ellis, Giele, Kunszt Melnikov:arXiv:0806.3467
We have calculated the
one-loop amplitudes for
ttgg and ttggg as a proof
of principle that the
method can be extended
to massive particles.
Calculation times are
longer than pure gluon
amplitudes
Thus ttgg ~ 10 ms (cf 1ms
for gggg) and ttggg ~ 40
ms.
W+qqggg amplitudes
Numerical stability assured
by computation, (where
necessary) in quadruple
precision.
'
0
=|(DP-QP)/QP|
Evaluation times are 45-50
msec per leading color
primitive on 2.33 GHz
pentium Xeon machine.
EGKMZ, arXiv:0810.2762
One-loop amplitude summary
There are a number of groups which use unitarity and
OPP ideas to perform one-loop calculations (Berger et al, OPP,
Lazopoulos, Giele & Winter).
The F90 Rocket program (Ellis, Melnikov, Zanderighi) can
compute results at one loop for:-
N gluon scattering amplitudes
two quarks (massless and massive) + N gluons,
W-boson + two quarks + N gluons,
W-boson + four quarks + 1 gluon
tt+N gluons, ttqq+ N gluons (EGKM +Schulze)
Note that extension to arbitrary number of gluons (using
Berends-Giele recursion), and the proven ability to deal
with massive fermions.
W + 3 jets
Here I report on recent calculations of W+3-jet rate
at hadron colliders
The calculation represents a proof of principle for
the unitarity-based methods and is challenging with
traditional methods, (1480 1-loop diagrams)
W+3 jets is phenomenologically relevant because
of Tevatron measurements, single top, SUSY
searches,
More generally the rates for vector boson + jets
production at the LHC are important as
backgrounds to BSM processes.
RKE, Melnikov,Zanderighi arXiv:0901.4101
W+3 jets: First NLO QCD results
We simplify the problem by working at large N
c
(N
c
is the number of colors) and by keeping only
the two quark channels qqW+ggg
These are 10-30 percent approximations, so the
phenomenology is rather preliminary.
Virtual corrections are computed using a grid
determined from the leading order computation
Dipole subtraction is used for the real emission
corrections.
RKE, Melnikov, Zanderighi, arXiv:0901.4101
W+3 jets: First NLO QCD results
for LHC
Inclusive W+3jets + K factor
p
T
>50GeV, |(| <3, R=0.7,
)S=14 TeV
This calculation displays the
standard improvement of scale
dependence.
Detailed phenomenology at the
10% level will have to await the
inclusion of all processes.
RKE, Melnikov,Zanderighi arXiv:0901.4101
Summary
!
S
(M
Z
) is known to < 1% and parton distributions are known
well enough to predict most cross sections to 20%,
(0.005<x<0.3).
Theoretical error in jet shapes in e
+
e
-
annihilation is now for the
first time smaller than the experimental error.
New analytic results on total top production at NLO. Drive to
complete NNLO calculation.
At high p
T
, parton level integrators, such as MCFM, can do an
adequate job of describing data with smaller theoretical errors
than other methods.
Summary (continued)
Open theoretical problem in calculating multi-leg
processes at NLO has been the calculation of one-loop
amplitudes
All one-loop integrals for QCD are known.
Unitarity based methods have achieved important
results for one-loop amplitudes, these methods are
now being tested in real physical calculations.
The hope is to have several semi-automatic methods
of calculating one-loop amplitudes.
Collaborators (1974-2009)
G. Altarelli, P. Arnold, R.D. Ball, J. M. Campbell, J.C. Collins,
S. Dawson, B.A. Dobrescu, F. Febres-Cordero, M.A. Furman,
W. Furmanski, H. Georgi, W.T. Giele, R.J. Gonsalves, J.F. Gunion,
M. Greco, H.E. Haber, F. Hautmann, I. Hinchliffe, J. Kalinowski,
Z. Kunszt, E.M. Levin, M. Machacek, L. Maiani, F. Maltoni,
G. Marchesini, G. Martinelli, K. Melnikov, P. Nason, G. Parisi,
S. J. Parke, R. Petronzio, S-Y. Pi, H.D. Politzer, C. Quigg,
D. L. Rainwater, L. Reina, M.H. Reno, D.A. Ross, G.G. Ross,
W.G. Scott, J.C. Sexton, M. Soldate, W.J. Stirling, A.E. Terrano,
F. Tramontano, J.J. van der Bij, S. Veseli, W. Vogelsang,
D. Wackeroth, B.R. Webber, S. Willenbrock, G. Zanderighi