Michael Beimforde

1
Investigations of the single tank pe spectrum
Investigations of the single tank pe spectrum

• Michael Beimforde
• Collaboration Meeting, October 6, 2006

2
Outline - Spectrum - Fits - TankToySim -
Conclusion

• Outline
• Single tank pe spectrum from calibration data
• Introducing fit to the calibration data
• Test of the fits robustness
• Simplified toy-simulations for geometrical
  analysis of the tank

IceCube Collaboration meeting 2006, Michael
Beimforde, 2/11
3
Outline - Spectrum - Fits - TankToySim -
Conclusion
Single Tank Spectrum
muon peak

• -gt weekly calibration run - data
• -gt recorded with single tank trigger
• .hit files read out using
• fatreader (-c --sumw2 --step20)
• domcal (with droop correction)
• topwaveprocessor (_at_43Ohm)
• flat-ntupel
• charge cut 40pe ... 500pe

background
not just exp
IceCube Collaboration meeting 2006, Michael
Beimforde, 3/11
4
Outline - Spectrum - Fits - TankToySim -
Conclusion
Fitting the spectrum

• Fit should be
• fast
• robust
• fitting the whole range

Landau exp exp
1 VEM
1 VEM

• Starting parameters extracted from
• number of entries
• the mean value
• p_2,p_4,p_6const
• p_0entries
• p_3,p_5log(entries)const
• p_1meanconst

IceCube Collaboration meeting 2006, Michael
Beimforde, 4/11
5
Outline - Spectrum - Fits - TankToySim -
Conclusion
Demo fits
IceCube Collaboration meeting 2006, Michael
Beimforde, 5/11
6
Outline - Spectrum - Fits - TankToySim -
Conclusion

• Testing fit robustness
• two different calibration runs (Run30833,
  Run30858)
• assuming that they are stable
• dividing them in several sub-runs
• fitting each dom for each sub-run
• testing for VEM-variations and ?2/ndf
  distribution

sample DOM VEM distribution
of entries
?
charge / pe
of entries
sample DOM ?2/ndf distribution
mean
????ndf
IceCube Collaboration meeting 2006, Michael
Beimforde, 6/11
7
Outline - Spectrum - Fits - TankToySim -
Conclusion
Results
all DOMs

• Mean standard deviation around lt3 pe
• Mean mean ?2/ndf slightly larger 1.0

• DOM 67-63 with problematic spectrum

??in pe
64-1
all DOMs
mean ?2/ndf
IceCube Collaboration meeting 2006, Michael
Beimforde, 7/11
8
Outline - Spectrum - Fits - TankToySim -
Conclusion

• TankToySim
• Easy MC for muon track length distribution
• Scaling this track length distribution to a VEM

Through top
Through side wall
IceCube Collaboration meeting 2006, Michael
Beimforde, 8/11
9
Outline - Spectrum - Fits - TankToySim -
Conclusion
1 VEM165.0 pe
through lid bottom
edge clipper
horizontal
IceCube Collaboration meeting 2006, Michael
Beimforde,9/11
10
Outline - Spectrum - Fits - TankToySim -
Conclusion
TankToySim
Detector response

• TankToySim spectrum

• tanktop VEM simulation Landau
  fit

• Convolving TankToySim with Landau

tanktop

• Similar to tanktop result but too complicated for
  calibration

IceCube Collaboration meeting 2006, Michael
Beimforde, 10/11
11
Outline - Spectrum - Fits - TankToySim -
Conclusion

• Conclusions
• A fit to the whole spectrum could be useful for
  easy calibration procedures
• Landau exp exp is a possible way to fit the
  spectrum
• Geometrical dependency of the spectrum was
  investigated

all DOMs Run 30833
station number
IceCube Collaboration meeting 2006, Michael
Beimforde, 11/11
12
Outline - Spectrum - Fits - TankToySim -
Conclusion
First try
Landau exp (lan1exp)
no distinction between short and long lived
exponential parts -gt results in fit problems at
high and low charges
IceCube Collaboration meeting 2006, Michael
Beimforde, 12/11
13
Outline - Spectrum - Fits - TankToySim -
Conclusion
Pulls
IceCube Collaboration meeting 2006, Michael
Beimforde, 13/11