sd energy

Absolute Energy Measurementby Ground Array:AGASA and Auger-SD: 

University of California, Los Angeles Department of Physics and Astronomy arisaka@physics.ucla.edu Katsushi Arisaka Absolute Energy Measurement by Ground Array: AGASA and Auger-SD

What is the Problem?: 

What is the Problem? AGASA sees a dozen of super GZK Events. But HiRes sees GZK cut-off? A factor of two difference in the Flux below GZK cutoff.

HiRes vs. AGASA: 

HiRes vs. AGASA Flux difference of the factor of two can be explain by energy difference of 20-30%. Is there any systematics error contributing to flux distortion above GZK cutoff?

What are we facing?: 

What are we facing? Is the Pierre-Auger going to uncover their mistakes and fix them? or Are we going to repeat the same mistakes? That is the question of systematics, not statistics.

Our Goal of Systematic Errors: 

Our Goal of Systematic Errors AGASA/HiRes claim that their systematic error in energy measurement is AGASA: ~18% HiRes: ~21% Their factor of two discrepancy in flux is fully consistent with their systematics. To solve this discrepancy in Auger, we should aim at Auger-SD:  10% Auger-FD:  10%

Outline: 

Outline Comparison of AGASA and Auger-SD from detector response point of view. Energy Estimate by AGASA Systematic Error Possible origin of “Over-linearity” Energy Estimate by Pierre-Auger Systematic Error

AGASA Detector: 

AGASA Detector 111 Electron Detectors 27 Muon Detectors Optical fiber cables Operation since1991

Pierre-Auger SD: 

Pierre-Auger SD 3m

AGASA vs. Auger-SD: 

AGASA vs. Auger-SD

Particle Distributions: 

Particle Distributions

Detector Response: 

Detector Response AGASA Auger-SD  e  Energy Deposit  e  1.2m Water 5cm Plastic

Detector Response: 

Detector Response AGASA Auger-SD  e  Energy Deposit  e  1.2m Water 5cm Plastic Acceptance  cos  Acceptance > cos 

AGASA vs. Auger-SD: 

AGASA vs. Auger-SD Both use VEM as one unit, but they are different. AGASA 10MeV equivalent Auger-SD 240MeV equivalent Easier to calibrate in Auger (24 times bigger signal than AGASA). For EAS, Auger is more sensitive to energy carried by muons. Gamma rays deposit more energy in Auger-SD. AGASA ~10% Auger-SD ~100% Auger is totally calorimetric for EM Showers. (AGASA is not.) Acceptance is more uniform in Auger-SD (as a faction of Zenith Angle). AGASA readout is Log-Amp, Auger is FADC.

Structure of Linear-focus PMT : 

Structure of Linear-focus PMT QE CE 1 2 3 n N G = 123  n

Definition: 

Definition QE(, ): Quantum Efficiency, as a function of Wave Length () and photon incident angle (). CE(B): Photoelectron Collection Efficiency, as a function of external magnetic field (B). G(HV,T,IA): PMT Gain as a function of High Voltage (HV), Temperature (T) and Anode Current (IA). dE/dX: Energy loss per unit length (in MeV/g/cm2). Opt: Photon acceptance of optical system. Abs(): Absorption Length of Water in Tank. Ref(): Effective Tyvek Reflectivity.

AGASA Energy Calibration: 

AGASA Energy Calibration Vertical Equivalent Muon (VEM): EAS signal on Scintillator:

Energy Deposit per Scintillator: 

Energy Deposit per Scintillator Visible energy deposit on Scintillator: Same Scintillator/PMT  Cancels out!

AGASA Energy Determination: 

AGASA Energy Determination

VEM spectrum at AGASA (one day): 

VEM spectrum at AGASA (one day) Detector Gain is monitored by muons in each run every day.

AGASA Muon peaks over 7 years data: 

AGASA Muon peaks over 7 years data Seasonal variation is clearly shown.

Systematic Error in AGASA: 

Systematic Error in AGASA Visible energy deposit on each Scintillator of AGASA: Same Scintillator/PMT  Cancels out! Non Linearity 7% VEM Calibration 5%  Total systematic error ~9%

AGASA Detector Response: 

AGASA Detector Response ICRC2001 Sakaki et al

2x1020eV event at AGASA: 

2x1020eV event at AGASA E=200EeV, Emin = 160EeV Be aware that the density is given in unit of 10MeV/m2.

AGASA Energy Estimate by S(600): 

AGASA Energy Estimate by S(600) ICRC2001 Sakaki et al S(600) is far more stable than Nch.

Slide25: 

Missing Energy at AGASA By M. Teshima Invisible energy is 5-10% of total energy. It depends on Composition and Initial interaction model.

Slide26: 

Old Conversion from S(600) to Total Energ by AGASA ICRC2001 Sakaki et al E=2.031017S0(600)1.00 eV

Modified Energy Conversion from S(600): 

Modified Energy Conversion from S(600) E=2.211017S0(600)1.03 eV

AGASA Composition Study: 

AGASA Composition Study Transition from Iron to Proton seen. For better energy estimate, composition needs to be understood. astro-ph/991222 Nagano et al by CORSIKA

AGASA Energy Resolution(Expected statistical fluctuation): 

AGASA Energy Resolution (Expected statistical fluctuation) /E=24% /E=21%

Concerns on Log-Amp Readout: 

Concerns on Log-Amp Readout AGASA measure the total energy deposit by integrating the charge and discharge it with time constant of =10sec. This is valid only if pulse width 100nsec. But this is not true because of Arrival time structure of shower front. Delayed Particles (such as slow neutrons).

Arrival time of Shower particles: 

Arrival time of Shower particles Arrival time of shower is given by Energy is over-estimated by =10sec

Systematic Error due to Shower Front Structure: 

Systematic Error due to Shower Front Structure According to this study, the energy overestimate is ~4% independent of the cosmic ray energy. However, if the correction depends on energy, it could introduce “over-linearity” which compensate “ZGK structure”.

AGASA managed to calcel out GZK structure?: 

AGASA managed to calcel out GZK structure?

Systematic Error due to Delayed Particles: 

Systematic Error due to Delayed Particles If there are ND delayed particles at tD with respect to Ni incident particles at t=0, the pulse height is Energy is over-estimated by =10sec

Systematic Error due to Delayed Particles: 

Systematic Error due to Delayed Particles According to this study, the energy overestimate is ~3% independent of the cosmic ray energy. However, if the correction depends on energy, it could introduce “over-linearity” which compensate “ZGK structure”.

Personal Remarks: 

Personal Remarks There is a slight change that the GZK cutoff exists in nature, but AGASA somehow managed to compensate it by artificial “over-linearity” caused by Energy-dependent arrival time of shower particles. Energy-dependent delayed particles (such as slow neutrons.) Pierre-Auger has so such artifacts as we use FADC instead of Log-Amp.

Summary of Systematics in Total Energy Measurement by AGASA: 

Total visible energy is given by fitting LDF. LDF Estimate  5% Attenuation on S(600) for inclined Shower  5% Shower front structure  5% Delayed particles  5% Then, the absolute energy of the primary cosmic ray is given by Etotal = Evis + Eneutrinos + Emuons Eneutrinos/ Etotal ~ 10  5% Emuons/ Etotal ~ 5  5% (depends on composition) For AGASA, muons are invisible. They took conservative approach of energy estimate (to obtain lower bound). These additional “non-detector” systematic error is ~15%. Summary of Systematics in Total Energy Measurement by AGASA

AGASA Systematics Estimate: 

AGASA Systematics Estimate

Systematic Error of AGASA: 

Systematic Error of AGASA

Pierre Auger SD Tank: 

Pierre Auger SD Tank

Auger-SD Energy Calibration: 

Auger-SD Energy Calibration Vertical Equivalent Muon (VEM): EAS signal on each tank:

Energy Deposit per Water Tank: 

Energy Deposit per Water Tank Visible energy deposit on Water Tank: Same Tanks/PMTs

Auger-SD: Simulated Muon Humps vs. VEM: 

Auger-SD: Simulated Muon Humps vs. VEM

Auger-SD:Real Data vs. MC Simulation: 

Auger-SD: Real Data vs. MC Simulation EA Real Data

Effect of Photo-statistics: 

Effect of Photo-statistics Three Sum PMTs Single PMT

Systematic Error in Auger-SD: 

Systematic Error in Auger-SD Visible energy deposit on each tank of Auger-SD: Same Tanks/PMTs Non Linearity 3% VEM Calibration 5%  Total systematic error ~6%

Principle of SD Absolute Energy Measurement: 

Principle of SD Absolute Energy Measurement Thanks to cosmic ray muons in nature, almost all of the systematics are canceled out; no need to measure QE, CE, Gain etc. Environmental effects can be monitored online; they are also canceled out automatically anyway. Systematic errors come only from the difference of pulse shape (height and length) between VEM and EAS.

Comparison between VEM and EAS: 

Comparison between VEM and EAS Detector related systematics are due to the difference between VEM and EAS.

Systematic Errors Due to Poor Detector Performance: 

Systematic Errors Due to Poor Detector Performance

Auger LDF Simulation by QGSJET: 

Auger LDF Simulation by QGSJET GAP Note 2002-073, P. Billoir et al

Zenith angle dependence of S(1000): 

Zenith angle dependence of S(1000) GAP Note 2002-075, P. Billoir Iron Proton CORSIKA

Other Systematics in Total Energy Measurement by SD (taken from AGASA): 

Total visible energy is given by fitting LDF. LDF Estimate  5% Attenuation on S(1000) for inclined shower  5% Then, the absolute energy of the primary cosmic ray is given by Etotal = Evis + Eneutrinos + Emuons Eneutrinos/ Etotal ~ 10  5% Emuons/ Etotal ~ 5  10% (depends on composition) It is important to estimate Emuons (i.e. primary composition) from the pulse shape. Once it’s done, additional error can be reduced to ~10%. Other Systematics in Total Energy Measurement by SD (taken from AGASA)

How can we reduce the Systematic Error due to Emuons ?: 

How can we reduce the Systematic Error due to Emuons ? Detailed comparison of real data and detector simulation to extract #muons in showers. Infill with muon counters. Muon Steel Plate Scintillator

Systematic Error of Ground Arrays: 

Systematic Error of Ground Arrays

Summary: 

Summary Thanks to muons, almost all of the “detector hardware related” systematics are canceled out. AGASA might have introduced “over-linearity” due to Log-Amp which canceled out “GZK-cutoff in nature”. Auger is free from such a problem. Extra care must be taken for the difference between VEM and EAS. Both AGASA and Auger-SD have ~18% of systematic uncertainties. Once muon composition of Auger-SD is well measured, systematics can be reduced to ~11%.