Xe analysis meeting Bartender Accidental pileup Nsum3

1
Xe analysis meetingBartender (Accidental
pileup)Nsum3

• R. Sawada
• 18/Jun/2007

2
Bartender - accidental pileup in signal region
Slide from the previous meeting to remind the
procedure

• Introduced rotation of Michel positrons for
  efficient simulation of accidental pileup
• It performs the simulation in 1.6 sec/event.
  100,000 accidental backups were generated.
• Procedure
• Distribute RD events and Michel positrons
  randomly according to specified event rate in
  2.5 micro sec.
• If total energy deposit in the calorimeter is
  smaller than a threshold, it repeats the first
  step until total energy deposit becomes larger
  than the threshold.
• Search the most energetic gamma ray. From time of
  the gamma ray, it searches a positron whose
  kinematics meets the trigger condition.
• Rotate all Michel decay sub-events in 2.5 micro
  sec time region so that the selected Michel
  positron comes around the most energetic gamma
  ray. The time differences between the gamma and
  the Michel positron are distributed randomly
  within specified time window.

3
Distribution of gamma pileups in accidental
backgrounds
Cut Two gamma ray pileup Vertical Energy of
the most energetic gamma in a event Horizontal
Time difference of two gamma rays
Type 1
Type 3
Type 2
With Michel rotation With kinematics
pre-selection Opening angle gt 173 degree Energy
deposit in calorimeter gt 40 MeV Momentum of
positron gt 47 MeV Time difference of positron
and gamma lt 750 psec Event selection 48 lt energy
deposit lt 54 MeV
Without Michel With kinematics pre-selection
(looser than trigger) Opening angle gt 160
degree Energy deposit in calorimeter gt 30 MeV
Momentum of positron gt 40 MeV Time difference of
positron and gamma lt 50 nsec Event selection 48
lt energy deposit lt 54 MeV
Events are classified to three types 1 RD
gamma Michel positron gamma (no contribution
to trigger) 2 RD gamma Michel positron
gamma (contributes to trigger) 3 RD gamma
RD positron
4
Bartender - accidental pileup in signal region
Almost same as the slide at the previous meeting

• Introduced rotation of Michel positrons for
  efficient simulation of accidental pileup
• It performs the simulation in 1.6 sec/event.
  100,000 accidental backups were generated.
• Procedure
• Distribute RD events and Michel positrons
  randomly according to specified event rate in
  2.5 micro sec.
• If total energy deposit in the calorimeter is
  smaller than a threshold, it repeats the first
  step until total energy deposit becomes larger
  than the threshold.
• Search the most energetic gamma ray. From time of
  the gamma ray, it searches a positron whose
  kinematics meets the trigger condition.
• Rotate all Michel decay sub-events in 2.5 micro
  sec time region so that the selected Michel
  positron comes around the most energetic gamma
  ray. The time differences between the gamma and
  the Michel positron are distributed randomly
  within specified time window.

Modified to choose a gamma ray randomly to
prevent bias
5
Nsum3 and dependence (undergoing)

• Comparison of nsum and nsum3
• Comparison of nsum2 and nsum3

6
Nsum and Nsum3
nsum/nsum3 w
nsum/nsum3 v
nsum/nsum3 u
7
nsum2 and nsum3
Nsum3
Nsum3
Nsum2
Nsum2
w gt 4cm
w gt 4cm
u
w
w
u
Nsum2
Nsum3
Nsum3 has less dependence on position than nsum2.
While too much correction is done in shallow
part. Probably due to two reasons (shower and
scintillation photon scattering)
w gt 4cm
w gt 4cm
v
v
8
Resolution before position dependence correction
w gt 4cm
w gt 4cm
Nsum2
Nsum3
FWHM 8.9 sigma 3.0
FWHM 8.5 sigma 3.0
No significant difference. Resolution after
position correction must be studied. There are
some room to improve Nsum3 (effective solid
angle, use center of shower instead of first
conversion point...)
9
End
10
Nsum3(total charge corrected with total PMT
coverage)
C Normalization factor ( constant ) O Total
PMT coverage viewed from reconstructed position N
Number of photons observed by a PMT
We could also take into account attenuation and
scattering. Detailed study of the result is not
done yet. If we increase corrections, several
parameters can affect the energy resolution (
position reconstruction, attenuation length
estimation, simulation settings...). It is not
clear how much we should do correction.
Philosophy of this calorimeter is measuring
energy without complicated correction.
Nsum
Nsum2
Nsum3