high energy physics group, national taiwan university

High Energy Physics Group, National Taiwan University

Yee Bob Hsiung National Taiwan Universityfor NuTel group

UHE UHE workshop workshopApril 23-26, 2006April 23-26, 2006

IHEP, BeijingIHEP, BeijingIntroductionFeasibility StudyDetector StatusConclusion

Status of NuTel - a Neutrino Telescope for Observing PeV from AGN

Cosmic Rays and Neutrinos - Back to 2002

Cosmic Ray Spectrum

Not Well Understood

CR + X e2e

UHECR + CMB N + GZK

~ 0.1 /EeV/year/km2/sr

GZK Firm!

~ 1.2 x 103 /PeV/year/km2/sr

AGN ?

AGN Jets, CRs and AGN Jets, CRs and Protons?

?

Window of Opportunity

Conventional Detector

UHECR Detector?

Earth SkimmingEarth Skimming

Telescope

Cross SectionCross Section ~ E~ E1.41.4

Earth Skimming + Mountain PenetratingCherenkov vs. fluorescence

Sensitive to

e: electron energy mostly absorbed in mountain : no extensive air shower

appearanceexperiment!

Three simulation stagesThree simulation stages1. Mountain simulation:

+N cross-section– inelasticity– energy loss of tau

2. Air shower simulation: Cerenkov photons decay mode – CORSIKA detailed air shower simulation vs. f

ast simulation3. Detector performance simulation

– light propagation + Q.E.– pixelization for triggers – reconstruction

1

2

3

2002-2004

inside mountaininside mountain

• SM CC +N cross-section• Inelasticity & energy loss are calculated by G.L. Lin, J.J. Tseng, T.W.

Yeh, F.F. Lee of NCTU• Range (distance when survival prob = e-1) are calculated by M.A. Huang

Tau fluxTau flux

• Tau flux:– Fast simulation: single interaction inside target– Full-scale transport eq.: Consider multiple interactions

...• Conversion efficiency:

– optimal thickness ~ several times of

– Energy loss decreases conversion efficiency

dE/dxNo dE/dx

Lateral profile of Cerenkov photonsLateral profile of Cerenkov photons for horizontal shower (CORSIKA ) for horizontal shower (CORSIKA )

• Similar profile for showers produced by e– and • Cerenkov ring distance ~ (L-Rmax)tan c

• Outside ring, photon density ~ exponential decay• Detector can trigger far away from Cerenkov ring

1018 eV

1016 eV

1014 eV

Photons numbers vs opening anglePhotons numbers vs opening angle

No atten.

Photon density

Opening angle (radian)

1 PeV showerShower core to

detector plane

30 km awaySerious drop

with attenuation Atten.=15km

Optics assumptions (back in 2003)Optics assumptions (back in 2003)• ASHRA-type Mirror + a simple correctio

n lens

• Multi-Anode Photomultiplier with 0.5o x 0.5o pixel span

• Light collection : 1 m2 aperture, 8o x 16o field of view,

over all 10% efficiency for γ→ p.e.

The Signal and Background PatternThe Signal and Background Pattern

Cherenkov: ns pulse, angular span ~ 1.5 degreesNight Sky Background (mean)

Measured at Lulin observatory: 2.0 x103 ph/ns/m2/sr

A magnitude 0 star gives 7.6 ph/m2/ns in (290,390) nm

Cosmic Ray background very smallCluster-based trigger algorithm

Random Background with NSB flux 1 km away from a 1 PeV e- shower

Trigger ConfigurationTrigger Configuration• Single Pixel Trigger: One pixel pass energy threshold H

• Duo Trigger: Two neighbouring pixels pass threshold H

• H-L Trigger: Two neighbouring pixels with one passes high threshold H

and the other one passes low threshold L• Sum Trigger: 1. (3x3) trigger cell

2. Central pixel pass high threshold H

Neighbour Npe Sum pass threshold A=n1+n2+…+n8

Night Sky Background:

• Npe Follows Poisson distribution: Prob(n;μ) = e-μ μn/n!, μ = <Npe> Φ tg A FOV εA εq ,

Φ = 200/ns/m2/sr A=1m2 FOV=0.5ox0.5oεA =0.5 εq =0.2 μ =0.039 tg=25ns ; =0.076 tg=50ns

Range<0.5km: Majority of photon arrives within 25 ns

Most of photons arrives within 50 ns

H L

H H

H

Hn2n1 n3

n4n5n6n7

n8

NSB Trigger RateNSB Trigger Rate

50nsSingle Pixel Trigger: H=6H-L Trigger : (H,L)=(6,1)Duo Trigger : H=4Sum Trigger1: (H,A)=(1,9)Sum Trigger2: (H,A)=(2,8)

25nsSingle Pixel Trigger: H=5H-L Trigger : (H,L)=(5,1)Duo Trigger : H=3Sum Trigger1: (H,A)=(1,7)Sum Trigger2: (H,A)=(2,6)

For 10 Hz order NSB trigger rate, the Trigger Configurations are:

8 Npe

10 Npe

N=107MC, (32x32)Pixels

Trigger Efficiency for Electron ShowerTrigger Efficiency for Electron Shower

Sum Trigger givesThe largest range

1.1km for trig=90%

Sum trigger are similar

Other Three triggersare similar

Conservative estimation is 200 γneeded

90%

Preliminary ReconstructionPreliminary Reconstruction• Reconstruction: Minimize 2 for x,y,,,

and E– Two Detectors Separated by ~ 100m

E, x, y, ,

1 2

N1, T1, x1 , y

1, 1, 1

N2, T2, x2 , y

2, 2, 2

Possibility for Reconstruction Possibility for Reconstruction

• Possible to Reconstruct Events – Angular Error within 1°– Energy Error ~ 40%– Reconstruction Efficiency > 90% if triggered

Acceptance DeterminationAcceptance DeterminationIntegration of efficiencies in phase space

Three independent methods for cross-checking

Results are consistent with each other!

Method Efficiency Integration Investigator

MIR Range determined fromSimulation

Monte Carlo Alfred

MIME Modelled Curve Monte Carlo Minzu

NISE Detailed Simulation Numerical Ping

All three got consistent results

Best FOV Best FOV Site Rate

(/yr)

Hualalai

88º - 96º

49º - 81º ♤ 0.71

Loa 90º - 98º

2º - 34º ♤ 0.85

Kea 92 º - 100 º

-134º - 102º ♥

1.10

♤ FOV centered on Kea♥ FOV centered on Hualalai

Note that FOV is 8o x 32 o

MIMEMIME

Loa

Big Island

contour plot

• Pick energy• Put detector on top of Loa• Pick position randomly on a 20

km by 20 km vertical plane located 25 km north of Loa

• Emit randomly in 60o cone• Trace the track to find the exit p

oint of • Find decay point• Assume e/π took away ½ of en

ergy and find the shower core position (air density 10-3 g/cm3)

• Make sure shower core is above 1.5 km cloud level

MIMEMIMEMake sure the pathway is clear between exit point and shower coreFind the angle and distance between shower core and detectorDetermine the number of photons in the solid angle covered by detector and apply attenuation effect (18 km attenuation length)Set the threshold at 200 ’s and check the shower core in the FOV (vertical -8o-0o and horizontal -4o- 12o)Event rate = 7.5 x 400 x π x 0.1(duty) x 0.8(BF) x eff The obtained rate for Loa is 0.46 per year

AcceptanceAcceptance● Mauna Loa watching

Mauna Kea● Higher energy

shower Larger trigger area, but longer decay length (50 km @ 1 EeV)

SensitivitySensitivity4.7 ×102Defined as reachable

upper limit of fluxAssume F(En) = F0 En-2

Assume no signal in 2 years of observationFeldman-Cousin method for upper limits: 2.44 signal eventsTheo1: ~ 0.5 events/year

Schematics of electronicsSchematics of electronics

Trigger

Hamamatsu8x8 MPMT

16-channelspreamplifier

DAQ

PMT Preamp.

32 – channels Data Collection Module in cPCI

10 bit x40 MHzADC

FADC

bufferRAM

Triggerdaisychain

cycleRAM

Signal-sharing plate

2 m cable

HV powersupply

Front-end electronics Inside cPCI (PXI) chassis

Multi-anode PMT (MAPMT) “H7546” of 8x8 pixels is used as photon-sensitive device


Trigger

Hamamatsu8x8 MPMT


DAQ

PMT Preamp.


10 bit x40 MHzADC

FADC

bufferRAM

Triggerdaisychain

cycleRAM


2 m cable

HV powersupply


Computer-controlled HV power supply “VHQ-202M” in VME is used for MAPMT, 2 channels/module, 1 channel supplies 4 MAPMT (256 pixels)

“SBS” PCIVME adapter


Trigger

Hamamatsu8x8 MPMT


DAQ

PMT Preamp.


10 bit x40 MHzADC

FADC

bufferRAM

Triggerdaisychain

cycleRAM


2 m cable

HV powersupply


“SBS” PCIVME adapter

Signal-sharing plate is used for increasing dynamic range of the system in factor of about 10-20 times


Trigger

Hamamatsu8x8 MPMT


DAQ

PMT Preamp.


10 bit x40 MHzADC

FADC

bufferRAM

Triggerdaisychain

cycleRAM


2 m cable

HV powersupply


16-channels charge sensitive preamplifier transforms charge into voltage for digitising by pipelined ADC


Trigger

Hamamatsu8x8 MPMT


DAQ

PMT Preamp.


10 bit x40 MHzADC

FADC

bufferRAM

Triggerdaisychain

cycleRAM


2 m cable

HV powersupply


4 preamplifier boards and signal- sharing plate are connected to one MAPMT

Holes for mechanical purposesin future


Trigger

Hamamatsu8x8 MPMT


DAQ

PMT Preamp.


10 bit x40 MHzADC

FADC

bufferRAM

Triggerdaisychain

cycleRAM


2 m cable

HV powersupply


32-channels Data Collection module in cPCI (PXI) processes signals from 32 channels, has Trigger logic on the module and memory of 256 ADC clocks. If one event is 8 clocks (200 ns), memory could keep up to 32 events.


Trigger

Hamamatsu8x8 MPMT


DAQ

PMT Preamp.


10 bit x40 MHzADC

FADC

bufferRAM

Triggerdaisychain

cycleRAM


2 m cable

HV powersupply


There will be 16 DCM boards (512 channels) inside one PXI chassisDCM

System (CPU) card

Active cPCI extender for debugging


Trigger

Hamamatsu8x8 MPMT


DAQ

PMT Preamp.


10 bit x40 MHzADC

FADC

bufferRAM

Triggerdaisychain

cycleRAM


2 m cable

HV powersupply


DAQ – in Linux, inside cPCI (PXI) CPU card

Some testsSome tests

Trigger


DAQ

Preamp.


10 bit x40 MHzADC

FADC

bufferRAM

cycleRAMC

Fromgenerator

A

CCable ~20mCable ~20mCable ~20mCable ~20m

A

B

B

C

CD

D

Double pulse (~100 ns difference)Double pulse (~100 ns difference)

Telescope parametersTelescope parameters

FOV: 16o horizontally 8o vertically

Image size at FP: 24cm(H) X 12cm(V)

Aperture: 1m2

Light guide:reduces image 3:1.8

Photo sensor:8X8 MAPMTs w.512 pixelsX2for 2 telescopes

Three-fresnel surface telescopeThree-fresnel surface telescope

0*: 0.8mm, 4*: 0.8mm, 8*: 1.6mm

File = fresnel13.otx Scale = 0.1260 Y/Z-view

793.7653 mm

Front side: fresnel1.1 m

Both sides: fresnel

---: 8o

---: 4o

---: 0o

A fresnel-edge loss: 17%

Spot image of 3-fresnel telescopeSpot image of 3-fresnel telescope

SPOT DIAGRAM File = fresnel13.otx 10.00000 mm

Wavelength(micron)

0.40470

In : 788Out: 0

In : 780Out: 0

In : 772Out: 0

0.0000 0.0000/ deg 0.0000 4.0000/ deg

0.0000 8.0000/ deg

Telescope w. two-identical fresnel Telescope w. two-identical fresnel lenslens

15:29:33

Error function = 0.508347E+06 Scale: 0.12 04-Apr-06

208.33 MM

Front side: fresnel1.1 m

Back side: fresnel

Aspheric lens0.4 m

Tow fresnel lens are identical---: 8o

---: 4o

---: 0o

A fresel-edge loss: 11%

Spot image of two-identical fresnel Spot image of two-identical fresnel telescopetelescope

SPOT DIAGRAM File = fr01 10.00000 mm

Wavelength(micron)

0.38000

In : 788Out: 0

In : 780Out: 0

In : 780Out: 0

0.0000 0.0000/ deg 0.0000 4.0000/ deg

0.0000 8.0000/ deg

Light guide simulationLight guide simulation

0

0.2

0.4

0.6

0.8

1

0 20 40 60 80

incident(deg)

T

20mm 30mm 35mm 40mm 50mm

Light guideLight guide

ConclusionConclusion• NuTel is an experiment dedicated to Earth skimmin

g / mountaing watching• The PeV cosmic rate is ~ 1 event/year• The cost is low: O(1) million US dollars to build it• Very good project for training students• However, lack of funding support in last two years• Ceased collaboration with ASHRA last year• Look for new funding support from university this s

pring, if succeed will start the optics this summer• Schedule: prototype deployment in ~2007-2008 for initial observation

People who worked on NuTel beforePeople who worked on NuTel before• Italy: IASF, CNR, Palermo

– N. La Barbera, O. Catalano,G. Cusumano, T. Mineo,B. Sacco

• France: Paris, France F. Vannucci, S. Bouaissi

• USA: Hawaii J.G. Learned

• Japan: ICRR M. Sasaki

• Taiwan: – NCTS/CosPA3– G.L. Lin, H. Athar

(Faculty)

NTUHEP/CosPA2PIs: W.S. Hou & Y.B. HsiungHardware Team:

K. Ueno (Optics)Y.K. Chi (Electronics)Y.S. Velikzhanin (Electronics)J.G.Shiu (DAQ)M.W.C. Lin (Technician)Master students

Simulation Team:M.Z. Wang (Faculty)P. Yeh (Faculty)C.C. Hsu (Ph.D. student)

master studentsNUU

M.A. Huang + students (Faculty)

From 2002-2005

high energy physics group, national taiwan university

Documents

passes high threshold

low threshold lsum trigger

3x3 trigger cell

hz order nsb trigger

mountain simulation

central pixel

threshold hh

sum trigger1