Experimental calibra/on of the ARA radio neutrino telescope with an electron beam in ice R. Gaior, A. Ishihara, T. Kuwabara, K. Mase , M. Relich, S. Ueyama, S. Yoshida for the ARA collaboraFon, M. Fukushima, D. Ikeda, J. N. MaLhews, H. Sagawa, T. Shibata, B. K. Shin and G. B. Thomson 14th, July, 2017, ICRC2017 K. Mase 1
■ The ARA calibra/on with the TA-ELS (ARAcalTA) Performed in January, 2015 at TA site, Utah � Purpose: BeLer understanding of the radio Vpol antenna � emissions and the detector calibraFon � Hpol antenna � We measured ² PolarizaFon ² Angular distribuFon Bicone ARA ² Coherence antenna LNA + filter (230-430 MHz) � 150-850 MHz � Ice target � Antenna tower Extendable: 2-12m � TA LINAC � 40 MeV electron beam line � 14th, July, 2017, ICRC2017 K. Mase 2
� Measured bunch structure ■ TA LINAC ~10 bunches � ü 40 MeV electron beam 2 ns � ü Typical electron number per bunch train: 2×10 8 → 30 PeV EM shower ü Pulse frequency: 2.86 GHz → pulse interval: 350 ps ü Bunch train width was opFmized to ~2 ns ü Beam lateral spread: ~4.5 cm ü Trigger signal available Cover wide range ü Electron number can be monitored (~3%) → Coherence � 2 × 10 8 electrons → Enough signal strength Correlation → Monitor of electron number � 14th, July, 2017, ICRC2017 K. Mase 3
■ Ice target and the configura/ons l 100 x 30 x 30 cm 3 Thermometer � observation l Easily rotatable structure emission angle � angle � l Easily movable on a rail l Plastic holder for the ice has a hole underneath for the beam ice inclination Dry ice (on side) � 1 m � angle ( α ) l Main data sets 40 MeV electrons � l With ice (30°, 45°, 60°) 40 MeV electron beam line � l No target � Ice target � Cherenkov angle in ice (56°) � α =30° � α =60° � Electrons stop after running 20 cm in ice due to an ionization loss R. Gaior, 1135, ICRC2015 � → Wide angle distribution emission angle [deg.] � 14th, July, 2017, ICRC2017 K. Mase 4
■ Expected radio emissions ü Several radio emissions are expected ü Askaryan radiation ü In ice ü Wide angular distribution due to the short tracks Sudden appearance ü Peak at more horizontal direction than the signal � Cherenkov angle (56°) ü Transition radiation Askaryan radiation � ü At air/ice boundary ü Peak at Cherenkov angle (56°) Transition radiation � ice � ü Sudden appearance radiation ü When beam appears ü Forward emission (Cherenkov angle is ~1°) ü More in Krijn’s talk on Tuesday 20 cm hole � 40 MeV electrons � metal � Electron Light beam pipe � Source facility � 14th, July, 2017, ICRC2017 K. Mase 5
■ Simula/on Bunch structure � Lateral distribution � � 2 ns � Electron beam (Geant4) � Including accelerator configurations � 4.5 cm � 20 cm � E-field calculation � Based on the classical EM theory (Lienard-Wiechert potentials) Middle point method (PRD 81, 123009 (2010)) Endpoints method (PRE84, 056602 (2011)) Ray trace � Thanks to Anne Zilles for sharing her code for the implementation tables made � Emission in ice � E-field � Obs. angle 0° (no target) at 1 m � ice � Emission in air � E-field [V/m] � Middle point Endpoints 20 cm hole � 40 MeV electrons � metal � Electron Light beam pipe � Source facility � 14th, July, 2017, ICRC2017 K. Mase 6
■ Detector simula/on R. Gaior, 1135, ICRC2015 � Obs. angle 0° (no target) at 1 m � Antenna response (T-domain) � E-field � filter + LNA response � Antenna height [m] � E-field [V/m] � Gain [dB] � 230-430 MHz � 5 ns � 2 ns � ＋� ＋� Time [ns] � Time [ns] � Frequency [GHz] � Voltage [V] � = � Time [s] � Verify the understanding the emission mechanisms and detector responses, comparing with data � 14th, July, 2017, ICRC2017 � 7� K. Mase �
� � ■ Comparisons of waveform and the frequency spectrum Voltage [V] 0.3 Simulation Simulation Vpol � Vpol � Data Data 0.2 0.1 0 -0.1 -0.2 -0.3 40 60 80 100 120 140 Time [ns] 0.3 Voltage [V] Configuration: Hpol � Simulation Ice 60°, obs. angle: 15° 0.2 Data 0.1 0 ü Reasonable agreements between data and simulation after correcting the cable -0.1 attenuation -0.2 ü Less Hpol signal → high polarization -0.3 ü Some indications of noise 40 60 80 100 120 140 Time [ns] 14th, July, 2017, ICRC2017 K. Mase 8
■ Cable / connector aYenua/on correc/on Faraday Cup (for the electron charge measurement) � TA short cables / connectors (~3m, up to 500 MHz) � Long cables (40m, high oscilloscope � frequency adapted) � 40 MeV electrons � Counting house � TA LINAC � Voltage [V] � Long (45 m) Charge loss Original 9.9 % Long (45 m) Short (3 m) 23.5% Short (3 m) Short + Long (40 m) � 45.7% � Short + Long (40 m) � Time [ns] � ² Found out the TA short cable attenuate signal significantly ² Electron number for data was underestimated by 46% ² The emission power is proportional to the charge square → correction of x2.1 (1.46 2 ) ² Original bunch structure turned out to be more narrower Group review � 9� K. Mase �
■ Polariza/on Time development of polarization � Configuration: Ice 60°, obs. angle 15°, Vpol � Polarization � Voltage [V] Voltage [V] � 0.3 Data Data (Ice target) 0.2 Simulation � Simulation (Ice target) 0.1 Data (No target) � 0 -0.1 -0.2 -0.3 40 60 80 100 120 140 Ice target 0.92±0.03 Time [ns] Time [ns] � SimulaFon 1.00±0.01 Polarization � No target 0.82±0.03 Data Highly Polarization angle � Simulation � polarized � Time [ns] � ü All signals shows relatively high vertical polarization ü Smaller polarization for the no target configuration ü Less polarized signals for the outside of the main peak → indication of the noise contamination 14th, July, 2017, ICRC2017 K. Mase 10
■ Coherence Time development of coherence � Configuration: Ice 30°, obs. angle 0°, Vpol � 0.3 Voltage [V] Voltage [V] � Data 0.2 -10.2 (radio energy [J]) Simulation � 0.1 Slope index: 1.86 ± 0.01 � -10.4 0 -10.6 -0.1 -0.2 -10.8 10 -0.3 Log 0 20 40 60 80 100 120 140 160 180 200 -11 Time [ns] Time [ns] � -11.2 -11.4 Data Slope index � -11.6 -11.8 7.6 7.8 8 8.2 8.4 8.6 Log (electron number) 10 ü High coherence, but not full Time [ns] � → Possibly due to the noise ü High coherence around the all main pulse ü Similar values for all the configurations (Even Hpol too) � ü Noises seem to show the high coherence � 14th, July, 2017, ICRC2017 K. Mase 11
■ Angular distribu/on distance corrected, but antenna gain not corrected � θ c (30°) � θ c (60°) � θ c (45°) � ice � Observation angle � 40 MeV electrons � Preliminary � Electron Light Source facility � ü Noise were filtered using a time window (±5 ns) with respect to the peak in a waveform ü Reasonable agreement after applying the noise cut (otherwise the shapes do not agree, data is ~60% higher) � 14th, July, 2017, ICRC2017 K. Mase 12
■ Summary l We have performed an experiment at Utah using the TA-ELS for the better understanding of radio emissions and the ARA antennas l Highly polarized and coherent signals were observed l Radio signals observed from the beam appearance l More signals observed when using an ice block l Agreements improved after correcting the cable attenuation l More detail simulation constructed to take the reflection and refraction into account l Need to understand the noise a little better l Like to understand how much Askaryan signals are contained using simulation 14th, July, 2017, ICRC2017 K. Mase 13
Backups � 14th, July, 2017, ICRC2017 K. Mase 14
■ Askaryan Radio Array (ARA) ² Aims to detect high energy neutrinos above 30 PeV using Askaryan radiaFon ² 37 staFons (3 staFons deployed so far) ² Each staFon has 4 strings at 200 m depth ² Each string has 2 Vpol + 2Hpol broadband antennas (~200–800 MHz) ² Total surface area ~100 km 2 ² ~10 IceCube @ 1 EeV � 2 km Astroparticle Physics 35 (2012) 457–477 � 14th, July, 2017, ICRC2017 K. Mase 15
■ Systema/c uncertain/es [dBi] � 1 ns � Original Long (45 m) Short (3 m) Short + Long (40 m) � Time [ns] � Angle [deg.] � Ratio � Bunch after cable correction Widest case Item Data Simula/on Shortest case � StaFsFcal error ±7% ±10% Stability ±19% - Antenna response -17% +14% uncertainty 1 ns � Bunch width - -14% +17% Sum ±20% ±24% Bunch number � Group review K. Mase 16
■ W aveform agreements (no target) Data Simulation � 0 m � 3.5 m � Δ t = 137.5 ns Δ t = 138.9 ns 7.4 m � ü Timing is matched with a cross correlation Δ t = 139.7 ns ü Time difference between data and simulation is explained by cable delay etc. ü Agreements are reasonable ü Prepulse seen � 17/07/14� 17�
■ Residual signal (no target) Data - Simulation � 0 m � 3.5 m � ü Similar shape 7.4 m � ü Similar amplitude ü Time correlates with signal → generated at source? Prior to the signal? ü 10 ns earlier (3 m below the cover box) ü From the end of the beam pipe? ü No distance dependence? 17/07/14� 18�
■ Spectrum agreements (no target) Data Simulation � 0 m � 3.5 m � 7.4 m � ü Similar behaviors ü Difference becomes larger at high height because the signal is weaker and the effect of additional component is larger � 17/07/14� 19�
■ Residual spectrum (no target) 0 m � 3.5 m � 7.4 m � ü Quite similar shape ü Flat spectrum → � White noise? � 17/07/14� 20�
Recommend
More recommend
