1. Description of the ATLAS jet veto measurement using the Banfi-Marchesini-Smye equation
Cyrille Marquet
Centre de Physique Théorique
Ecole Polytechnique
in collaboration with
Y. Hatta, C.Royon, G. Soyez, T. Ueda and D. Werder
Phys. Rev. D87 (2013) , arXiv:

2. The jet-veto measurement
ATLAS Collaboration, 2011
GAP FRACTION :
dijets events surviving veto / total dijets events
two high-pT jets with cone size R
large rapidity separation Δy
veto on additional jet activity with kT > Q0 in solid angle COUT
Q0 >> ΛQCD

3. Comparison to QCD calculations
regime of interest for this talk:
pT >> Q0
the standard NLO + parton shower approach
fails to describe the data at large Δy
resumming BFKL type logarithm
(HEJ approach) does not help
Andersen and Smillie (2011)
resummation of emissions of soft gluons at large angle needed

4. The fixed-order approach
expansion of the cross section ratio
Soyez (2011)
Sudakov type behavior of the form

5. Gluon emission at large angles
insensitive to the collinear singularity
not taken into account in parton showers
resummation of soft emissions
when pT >> EOUT, one can
resum the soft logarithms
while requiring that the energy flow
into the inter-jet region is less than EOUT
there are 2 types of soft logarithms: Sudakov logs and non-global logs
caveat: we identify EOUT with Q0
but in principle they are different

6. Sudakov logarithms emissions from primary partons
e.g. Oderda and Sterman (1998)
emissions from primary partons
- real emissions allowed for kT < EOUT
- virtual emissions allowed for kT < pT
miscancellation between the
real and virtual contributions
large logs
the resummation exponentiates

7. Non-global logarithms
Dasgupta and Salam (2001)
emissions from secondary partons
one should also forbid secondary emissions into the interjet region
parametrically of the same
order as the Sudakov logs,
but not easy to resum
(no exponentiation)
the multi-gluon configuration in the
inter-jet region is complicated
simplifications in the large Nc limit
used to derive
the BMS equation

8. Banfi-Marchesini-Smye equation
Banfi, Marchesini and Smye (2002)
probability Pτ that the total energy
emitted outside of the cones is less than Eout
differential probability for
the soft gluon emission
Sudakov logs
non-global logs
resummation variable
“out”
“in”
analytical insight and numerical solutions are available
Hatta and Ueda (2009)

9. Jet veto cross section in p+p
the BMS equation can be generalized to hadron-hadron collisions
schematically, at LO the gap fraction is given by:
GAP FRACTION =
example channels:
the different color structures impose different powers of the BMS probability
this is a bit more complicated, not all
powers of Pτ have the same arguments,
and also the different possible color
connections give different contributions
draw the diagrams in the large Nc limit,
the number of lines is the power of Pτ

10. Comparison to ATLAS data I
possible selections of the di-jet system: “leading pT” and “forward/backward”
caveat: our LO calculation does not allow to distinguish the two cases
pT dependence
good description
of the forward/backward data
leading-pT data
at the edge of
uncertainty band

11. Comparison to ATLAS data II
Δy dependence
good description
of the forward/backward data
leading-pT data
at the edge of
uncertainty band
Q0 dependence
we do not describe the Q0 dependence, this is expected since for pT ~ Q0
the soft gluon resummation is not important and we are missing NLO terms

12. Conclusions a new QCD description of the ATLAS jet-veto measurement
- for pT >> Q0 , standard NLO + parton shower doesn’t work
- instead: LO+resummation of soft gluon emissions at large angles
- this is done using the BMS equation with running coupling
can one extend our calculation to NLO ?
- i.e. fold the BMS probability on top of 2-to-3 partonic processes ?
- not clear that this is feasible
- if yes, we could attempt to describe both data sets
- we should in principle also get a better data description at large Q0
intermediate solution ?
- matching of LO+resummation approach to NLO