MC simulation
To obtain robust estimates of the signal acceptance and the background level, extensive stud-
ies have been performed to produce and validate Monte-Carlo (MC) samples. For the main
background process, the decay K+ → π+ π+ π− with two secondary decays of pions, a ded-
icated MC sample has been produced with the event preselection at the generator level. A
special approach has been developed for the signal MC production: a multiple mass sample
has been generated to cover the range of kinematically allowed scalar masses with a fixed
mass step and scalar lifetime τ0 . The event reweighting procedure is used to obtain signal
MC samples with various lifetimes τ different from τ0 .
Particle identification
The particle identification (PID) is based on the information from the electromagnetic
calorimeter LKr, the STRAW spectrometer and the muon detector MUV3. In particular,
the cut on E/p value is applied (E is the energy deposit in LKr associated with the track,
p is the track momentum measured by the STRAW) supplemented by the requirement of at
least one associated hit in MUV3 for a muon and the absence of such hits for a pion. The
misidentification (misID) probability for a pion to be identified as muon has been measured
as a function of p. This probability is used for the data-driven estimate of the backgrounds
involving one or two misID, e.g. K+ → π+ π+ π− with one secondary pion decay and one pion
misidentification. The RICH detector is not used for the PID, since it has been found to be inefficient for
multi-track decays. However, the track time is measured by the RICH and its performance has been studied in details
and published in peer-reviewed articles.
Trigger performance
The choice of the normalisation channel with the same signature allows to use the same
trigger stream for the signal and normalisation and thus avoid uncertainties of the trigger
efficiency measurement in the final result at first order. However, trigger efficiencies have
been measured for the signal and normalisation decay mode by using dedicated emulators.
The emulator performance has been evaluated on the control data.
Data analysis
The scalar search is performed on the full NA62 data sample collected in 2016–2018. The
blind analysis procedure is adopted for the search. The backgrounds are evaluated both from
MC and data. Control regions are defined to validate the background estimation procedure.
Expected upper limits on B are calculated using the bayesian approach for the statistical
treatment.
Website of the project
No website has been developed for the project.
