1. Stato della Calibrazione EM
D.Banfi, L.Carminati, L.Mandelli

2. Sommario
Necessita’ di una calibrazione piu’ raffinata per i fotoni, che distingua fra fotoni convertiti e non convertiti
Studiati tre possibili approcci al problema:
Peso globale
Parametri indipendenti dal raggio di conversione
Parametri dipendenti dal raggio di conversione
Studio sostanzialmente concluso (solo alcuni dettagli rimangono da discutere)
Nota in preparazione sulla calibrazione dei fotoni

3. Energia depositata davanti al calorimetro
Efront
Photons
E=100GeV
0.7<η<0.9
L’energia depositata davanti al calorimetro varia sensibilmente con il raggio di conversione del fotone
not converted
all converted
Efront
Photons
E=100GeV
0.7<η<0.9
all converted
conv radius (from MC truth) <40cm
40cm < conv radius < 80cm
80cm < conv radius < 120cm

4. Energia depositata fuori dal cluster
Photons
E=100GeV
0.7<η<0.9
Anche l’energia depositata fuori dal cluster e’ dipendente dal raggio di conversione del fotone
-not converted
-all converted
all converted
conv radius (from MC truth) <40cm
40cm < conv radius < 80cm
80cm < conv radius < 120cm
Photons
E=100GeV
0.7<η<0.9

5. Photons 25GeV
All photons
Unconverted photons
Converted photons
ΔE = = 1.04 GeV
ΔE = – 24.94= 0.01 GeV

6. Photons 100 GeV
All photons
Unconverted photons
Converted photons
ΔE = = 0.81 GeV
ΔE = = 0.22 GeV

7. Samples and statistics
Photons
Energy
dataset
events
10 GeV
007081
~1.1M/1.2M
25 GeV
007082
50 GeV
007062
75 GeV
007083
~0.9M/1.2M
100 GeV
007063
200 GeV
007084
500 GeV
007085
~400K/400k
1 TeV
007086
~200/200k
Tot
~7.0M/7.8M
Considering the conversion probability (from 10 to 50% as a function of η) and the IsEm efficiency (from 40 to 90% as function of η and E) we have, in the region with more material in front of calorimeter, less than 1k converted photons for each cell and energy.
The parameter dedicated to converted photons are computed cell by cell adding up the events of 3 adjacent cells. For conversion radius studies the statistic from 5 cells is added.
All the shown distribution for Ereco refers to the stats of 3 cell, each one calibrated with its own parameters.

8. Dedicated parameters: η = 0.3
25 GeV
Eta = 0.3
Converted photos calibrated with:
‘’Standard’’ parameters (any distintion between conv and unconv)
parameters ‘’dedicated’’ to conv
50 GeV
Better linearity:
24.524.9
49.649.9
99.6100
Sigma of the fitted gaussian are compatible inside errors
100 GeV

9. Dedicated parameters: η = 1.1
25 GeV
Eta = 1.1
Converted photos calibrated with:
‘’Standard’’ parameters (any distintion between conv and unconv)
parameters ‘’dedicated’’ to conv
50 GeV
Better linearity:
23.524.7
48.749.9
98.8100
Sigma compatible inside errors
100 GeV

10. Linearity and resolution: Barrel Region
Improved linearity but almost the same resolution
Linearity
Eta = 0.3
Linearity
Resolution
Eta = 1.1
Resolution
Converted photos calibrated with:
Standard parameters
Dedicated parameters

11. Conversion radius dependance: η = 0.3
Conv – no conv radius eta 0.3
25 GeV
The conversion radius range (from MC truth) is dived into 4 bins in barrel (020,2040, 4060, 6080 cm) and 3 bins in the EndCaps (08,816,1680 cm): for each bin we compute a set of parameters
50 GeV
Eta = 0.3
Converted photons calibrated with parameter dedicated to converted:
independent from conversion radius
dependent from conversion radius
100 GeV
Linearity and sigma are compatible inside the errors.
2-5 % reduction of RMS

12. Conversion radius dependence: η = 1.1
Conv – no conv radius eta 1.1
25 GeV
The parameters for converted photons are computed in 4 different bins of conversion radius.
Eta = 1.1
Converted photons calibrated with parameter dedicated to converted:
independent from conversion radius
dependent from conversion radius
50 GeV
Linearity and sigma are compatible inside the errors. The RMS shows ~10% improvement with parameters dependent on conversion radius, due to the better calibration of low energy tails
100 GeV

13. Conclusions and perspectives
Conv – no conv radius eta 1.1
Parameters dedicated to converted photons, independent from cvradius, introduce:
improvement in linearity, with respect to the standard parameters, both in Barrel and EndCap
Negligible effect on resolution on all the barrel, small improvements in the EndCap (we need to extend the study to all the EndCap cells)
Parameters dependent on conversion radius give almost the same performance in term of linearity and resolution that the independent from cvradius ones, but reduce the low energy tails (~10 % of RMS reduction at high eta) in the Barrel. More studies are ongoing in the EndCap region.
Possible scenarios, in increasing order of complexity:
Overall weigth on energy:
Pro: very easy approach
Cons: effect only on linearity, we loose partly the meaning of the CalibHits approach
Dedicated parameters independent from cvradius
Pro: it’s is a “full” CalibHits approach
Cons: introduce a degree of complication in the calibration (another set of coefs)
Dedicated parameters dependent from cvradius
Pro: Better performance in the tail calibrations, almost the same performance of point 2 in term of linearity and resolution
Cons: Very difficult to extract, big complication in the calibration code
complexity