Orateur
M.
Mikhail Danilov
(ITEP-Moscow)
Description
The CALICE collaboration has developed highly granular calorimeter prototypes to
evaluate technologies for experiments at a future lepton collider, and evaluated
their performance in test beams. One important use of these data is the
validation of the physics models in GEANT4, especially those related to hadronic
showers. This validation is crucial if Monte Carlo simulations are to be used
to optimise the design of detectors for ILC or CLIC.
In this talk we discuss several features of hadronic showers recorded in the
CALICE Si-W ECAL and scintillator-tile HCAL. The high spatial resolution of the
calorimeters permits the investigation of shower shapes in unprecedented detail.
For example, the start point of the shower can be identified with high
precision, and the longitudinal development of the shower after this point
permits some discrimination between the various components of the shower
(nuclear spallation, electromagnetic components, MIP-like hadrons etc.).
Detailed substructure within the shower, such as track segments, can also be
identified. Many of these features present new challenges to the simulation
models. We also report on recent tests of the time-structure of hadronic
showers in a tungsten calorimeter - a topic of especial interest for a CLIC
detector where accurate time stamping is required.
The calorimeters being built and tested by the CALICE collaboration are intended
to be optimised for particle flow (PFA) reconstruction of jets at a future linear
collider. This places an emphasis on measuring showers with a high spatial
granularity, in order that nearby showers can be disentangled. The calorimeters
are generally non-compensating, in other words they do not give the same response to
photons as charged hadrons. However, the high level of detail recorded within
the showers can be exploited in the form of "software compensation". A variety
of techniques can be used to discriminate between the different components
within showers and hence to weight them differently so as to improve the energy
resolution. This approach is also found to improve the linearity of the
response. We report on several studies along these lines.
The CALICE data on single pion showers can also be exploited to validate the
particle flow concept. By superimposing the data from two separate showers,
removing the incoming particle track from one of them, we can emulate a neutral
hadron shower in the neighbourhood of a charged particle, at various separation.
By offering these hybrid events to the standard PandoraPFA algorithm we can
evaluate its ability to reconstruct neutral energy in a difficult
environment. We show the results of this procedure, in comparison with the
Monte Carlo models usually used for the testing of PFA codes.
Auteur principal
M.
Mikhail Danilov
(ITEP-Moscow)
