5.7 Supersymmetry
199
scalar sum of large-R jet masses M
J [1176, 1177], which is large for signal events
and small for SM production of light quark and gluon jets. Even if the signal does not
result in boosted heavy particles, M
J can still be a good discriminator in scenarios
where the jet multiplicity is very large, such that on average jets accidentally overlap
more often than in SM processes [1178]. The advantage of M
J over other discriminating observables is that it is largely uncorrelated to kinematic quantities such as
H T or p
miss
T , which are frequently used in SUSY searches. It is an excellent quantity
to define signal and control regions. An ATLAS search using 36.1 fb
−1 of 13 TeV
data [1179] is optimised for RPV SUSY scenarios with six to ten quarks produced
at tree level in the SUSY decay cascade. Trimmed large-R jets are used to select
events, where four or five jets with p T > 200 GeV are required in the signal regions.
Depending on the decay scenario, the value of M
J has to be larger than 0.6, 0.8 or
1 TeV. The background is obtained from deriving templates in the jet mass distribution of large-R jets in signal-depleted control regions. For each jet, a probability
density function is derived, which gives the relative probability for a jet with given
p T and η to have a certain mass [1180]. For each jet in the signal region, a randomised
value for the jet mass is generated from the probability density function, in order to to
derive the background estimate. The procedure is verified in validation regions and
the corresponding uncertainty is obtained from dedicated side bands with smaller jet
multiplicities. The analysis excludes gluino masses between 1000 and 1875 GeV in
gluino cascade decays with ten quarks, and improves upon previous results based
on 8 TeV data by more than 800 GeV [1180]. A recent analysis by ATLAS in the
all-hadronic final state, based on 139 fb
−1 of 13 TeV data [1156], targets cascade
decays involving W and Z bosons, gluino-mediated four-top production and RPV
cascades involving top quarks. The analysis uses PF jets and the corresponding p
miss
T
significance, which improves the separation between events where p
miss
T
originates
from detector effects or from weakly interacting particles with high p T . Large-R
jets are reclustered from small-R jets, where small-R jets originating from pileup
interactions are removed using the jet-vertex-tagger [442] prior to the reclustering.
Ungroomed large-R jets with p T > 100 GeV are then used to calculate M
J . The
final results are based on analysing the measured distributions in p
miss
T
significance.
The gluino mass limits for top squark-mediated gluino decays with RPV couplings
are 1.5 TeV for ˜
t masses between 400 GeV and 1.1 TeV. The variable M
J is useful in +jets final states as well, as demonstrated in a CMS search using 137 fb
−1 of
13 TeV data [1181]. Similar as in ATLAS, a reclustering of small-R jets is employed,
but with R = 1.4. Because of the large value of R, ISR radiation can lead to large
tails in M
J , such that a reliable background model needs to be derived from control
regions in data. This is achieved by a method exploiting that M
J is uncorrelated
from the transverse mass m T , calculated between the lepton and p
miss
T . The plane in
M
J versus m T is used to define three background-dominated regions, from which
the background is predicted in the signal region. The analysis results in excluded
gluino masses below 2150 GeV for ˜
χ
0
1 masses up to 700 GeV, in gluino-mediated ˜
t
production.
Another example of SUSY scenarios where jet substructure plays an important
role are models where the next-to-lightest SUSY particle is a neutralino ˜
χ
0
2 , with
199
scalar sum of large-R jet masses M
J [1176, 1177], which is large for signal events
and small for SM production of light quark and gluon jets. Even if the signal does not
result in boosted heavy particles, M
J can still be a good discriminator in scenarios
where the jet multiplicity is very large, such that on average jets accidentally overlap
more often than in SM processes [1178]. The advantage of M
J over other discriminating observables is that it is largely uncorrelated to kinematic quantities such as
H T or p
miss
T , which are frequently used in SUSY searches. It is an excellent quantity
to define signal and control regions. An ATLAS search using 36.1 fb
−1 of 13 TeV
data [1179] is optimised for RPV SUSY scenarios with six to ten quarks produced
at tree level in the SUSY decay cascade. Trimmed large-R jets are used to select
events, where four or five jets with p T > 200 GeV are required in the signal regions.
Depending on the decay scenario, the value of M
J has to be larger than 0.6, 0.8 or
1 TeV. The background is obtained from deriving templates in the jet mass distribution of large-R jets in signal-depleted control regions. For each jet, a probability
density function is derived, which gives the relative probability for a jet with given
p T and η to have a certain mass [1180]. For each jet in the signal region, a randomised
value for the jet mass is generated from the probability density function, in order to to
derive the background estimate. The procedure is verified in validation regions and
the corresponding uncertainty is obtained from dedicated side bands with smaller jet
multiplicities. The analysis excludes gluino masses between 1000 and 1875 GeV in
gluino cascade decays with ten quarks, and improves upon previous results based
on 8 TeV data by more than 800 GeV [1180]. A recent analysis by ATLAS in the
all-hadronic final state, based on 139 fb
−1 of 13 TeV data [1156], targets cascade
decays involving W and Z bosons, gluino-mediated four-top production and RPV
cascades involving top quarks. The analysis uses PF jets and the corresponding p
miss
T
significance, which improves the separation between events where p
miss
T
originates
from detector effects or from weakly interacting particles with high p T . Large-R
jets are reclustered from small-R jets, where small-R jets originating from pileup
interactions are removed using the jet-vertex-tagger [442] prior to the reclustering.
Ungroomed large-R jets with p T > 100 GeV are then used to calculate M
J . The
final results are based on analysing the measured distributions in p
miss
T
significance.
The gluino mass limits for top squark-mediated gluino decays with RPV couplings
are 1.5 TeV for ˜
t masses between 400 GeV and 1.1 TeV. The variable M
J is useful in +jets final states as well, as demonstrated in a CMS search using 137 fb
−1 of
13 TeV data [1181]. Similar as in ATLAS, a reclustering of small-R jets is employed,
but with R = 1.4. Because of the large value of R, ISR radiation can lead to large
tails in M
J , such that a reliable background model needs to be derived from control
regions in data. This is achieved by a method exploiting that M
J is uncorrelated
from the transverse mass m T , calculated between the lepton and p
miss
T . The plane in
M
J versus m T is used to define three background-dominated regions, from which
the background is predicted in the signal region. The analysis results in excluded
gluino masses below 2150 GeV for ˜
χ
0
1 masses up to 700 GeV, in gluino-mediated ˜
t
production.
Another example of SUSY scenarios where jet substructure plays an important
role are models where the next-to-lightest SUSY particle is a neutralino ˜
χ
0
2 , with