180
5 Direct Searches for New Physics
momentum due to the undetected dark matter particles. The p
miss
T
induced by the χ ¯
χ
system is balanced by a SM particle, which seemingly has been produced singly.
Hence the common notation Mono-X searches, where X can be any object identified
in the detector, e.g. a jet, γ , W , Z , H , …
This results in a special importance of p
miss
T
in dark matter searches, since any
signal is expected to have a considerably different p
miss
T
spectrum than irreducible
SM backgrounds, for example from Z → νν production. These backgrounds are
difficult to model to a sufficiently high accuracy, such that these are usually validated
and constrained by data in dedicated control regions. Care has to be taken that the
control regions of a search in a given channel do not overlap with signal regions
of other channels. If possible, control regions should be chosen with small p
miss
T ,
for example, a Z → +jets selection can be used to estimate the background from
Z → νν+jets in mono-jet searches. More difficult are backgrounds where events
are only partially reconstructed, for example W → ν with the lepton failing the
selection requirements or produced outside the acceptance. In addition, there are
events with spurious p
miss
T
from detector noise, beam halos or cosmic muons. These
need to be rejected by dedicated algorithms, and their efficiency and false-positive
rates have to be studied.
A large number of mono-X searches has been carried out by the ATLAS and
CMS Collaborations since the start of the LHC. Not for all signatures explored jet
substructure techniques are applicable or useful. However, in some channels these are
indispensable tools, providing a large increase in sensitivity or even enabling these
analyses. In general, boosted topologies are most relevant for m χ 100 GeV, where
the recoil of the χ ¯
χ system receives a large momentum. For masses greater than
about 100 GeV, boosted final states loose sensitivity and should be complemented
by resolved topologies where possible.
The first searches for dark matter have been mono-jet searches using 7 TeV data
by ATLAS [1029, 1030] and CMS [1031]. In these searches, a jet from initial state
radiation recoils against the χ ¯
χ system, resulting in a p
miss
T +jet signature. These
searches have sensitivity to a large number of BSM models, where dark matter particles are predicted to couple to quarks and gluons. When considering the radiation of
a photon instead of a jet from the initial-state quarks, the reduction of the signal cross
section due to the electromagnetic interaction instead of the strong force is compensated by much smaller SM backgrounds from Z γ and W γ production. Analyses in
the mono-γ channel by ATLAS [1032] and CMS [1033] using 7 TeV data achieve
sensitivities to couplings and masses comparable to the mono-jet searches.
The four times larger data set and increased
√
s results in improved constraints
from mono-jet and mono-γ searches at 8 TeV [1034–1037]. For m χ between 1 and
400 GeV, where the validity of the EFT approach is still guaranteed, upper limits on
the χ -nucleon cross section could be improved by a factor of about three relative to
the 7 TeV results. New channels have been investigated for the first time using 8 TeV
data. Searches for mono-W and mono-Z production have been carried out. Naively,
the radiation of a W boson from the initial state quark would not result in better
sensitivity relative to gluon, quark or photon radiation. However, if there is constructive interference between the diagrams where the boson is radiated from the u or d
5 Direct Searches for New Physics
momentum due to the undetected dark matter particles. The p
miss
T
induced by the χ ¯
χ
system is balanced by a SM particle, which seemingly has been produced singly.
Hence the common notation Mono-X searches, where X can be any object identified
in the detector, e.g. a jet, γ , W , Z , H , …
This results in a special importance of p
miss
T
in dark matter searches, since any
signal is expected to have a considerably different p
miss
T
spectrum than irreducible
SM backgrounds, for example from Z → νν production. These backgrounds are
difficult to model to a sufficiently high accuracy, such that these are usually validated
and constrained by data in dedicated control regions. Care has to be taken that the
control regions of a search in a given channel do not overlap with signal regions
of other channels. If possible, control regions should be chosen with small p
miss
T ,
for example, a Z → +jets selection can be used to estimate the background from
Z → νν+jets in mono-jet searches. More difficult are backgrounds where events
are only partially reconstructed, for example W → ν with the lepton failing the
selection requirements or produced outside the acceptance. In addition, there are
events with spurious p
miss
T
from detector noise, beam halos or cosmic muons. These
need to be rejected by dedicated algorithms, and their efficiency and false-positive
rates have to be studied.
A large number of mono-X searches has been carried out by the ATLAS and
CMS Collaborations since the start of the LHC. Not for all signatures explored jet
substructure techniques are applicable or useful. However, in some channels these are
indispensable tools, providing a large increase in sensitivity or even enabling these
analyses. In general, boosted topologies are most relevant for m χ 100 GeV, where
the recoil of the χ ¯
χ system receives a large momentum. For masses greater than
about 100 GeV, boosted final states loose sensitivity and should be complemented
by resolved topologies where possible.
The first searches for dark matter have been mono-jet searches using 7 TeV data
by ATLAS [1029, 1030] and CMS [1031]. In these searches, a jet from initial state
radiation recoils against the χ ¯
χ system, resulting in a p
miss
T +jet signature. These
searches have sensitivity to a large number of BSM models, where dark matter particles are predicted to couple to quarks and gluons. When considering the radiation of
a photon instead of a jet from the initial-state quarks, the reduction of the signal cross
section due to the electromagnetic interaction instead of the strong force is compensated by much smaller SM backgrounds from Z γ and W γ production. Analyses in
the mono-γ channel by ATLAS [1032] and CMS [1033] using 7 TeV data achieve
sensitivities to couplings and masses comparable to the mono-jet searches.
The four times larger data set and increased
√
s results in improved constraints
from mono-jet and mono-γ searches at 8 TeV [1034–1037]. For m χ between 1 and
400 GeV, where the validity of the EFT approach is still guaranteed, upper limits on
the χ -nucleon cross section could be improved by a factor of about three relative to
the 7 TeV results. New channels have been investigated for the first time using 8 TeV
data. Searches for mono-W and mono-Z production have been carried out. Naively,
the radiation of a W boson from the initial state quark would not result in better
sensitivity relative to gluon, quark or photon radiation. However, if there is constructive interference between the diagrams where the boson is radiated from the u or d