NEUTRINO POINT-SOURCE ANALYSIS IN ICECUBE Juan Antonio Aguilar - - PowerPoint PPT Presentation

neutrino point source analysis in icecube
SMART_READER_LITE
LIVE PREVIEW

NEUTRINO POINT-SOURCE ANALYSIS IN ICECUBE Juan Antonio Aguilar - - PowerPoint PPT Presentation

NEUTRINO POINT-SOURCE ANALYSIS IN ICECUBE Juan Antonio Aguilar Vulcano Italy, 2010 IC40 All-Sky Stacking Conclusions 1 IC40 CONFIGURATION 2 ICECUBE IceCube Completion with 80 (+ 6) strings by January 2011 IceCube 79 (2010) 79


slide-1
SLIDE 1

Vulcano Italy, 2010

1

IC40 All-Sky Stacking Conclusions

NEUTRINO POINT-SOURCE ANALYSIS IN ICECUBE

Juan Antonio Aguilar

slide-2
SLIDE 2

2

IC40 CONFIGURATION

slide-3
SLIDE 3

Juan Antonio Aguilar Vulcano Workshop 2010

ICECUBE

IceCube 59 (2009-10) IceCube 40 (2008-9) IceCube IC59 data is being

  • processed. Analysis starting

now. IC40 has approx. 2 x effective area of the previous

  • IC22. IC40 data has been

analyzed for point sources. Completion with 80 (+ 6) strings by January 2011 IceCube

3

IceCube 79 (2010) 79 strings are in operation.

IC40 All-Sky Stacking Conclusions

slide-4
SLIDE 4

Juan Antonio Aguilar Vulcano Workshop 2010

IC40 MUON SAMPLE

L3 (muon sample) Trigger L2 (muon filter)

Strings Year Livetime SMT rate (Hz) µ filter rate (Hz)

  • atm. ν

final rate 9 2006 137 d 80 6 1.7 22 2007 276 d 450 20 18/d 40 2008 375.5 d 1100 23 40/d 59 2009 320 d 1900 24 79 2010 2300 40

4

IC40 rate evolution Detector rates:

  • Atm. neutrinos
  • Atm. muons

IC40 All-Sky Stacking Conclusions Upgoing Downgoing

slide-5
SLIDE 5

Juan Antonio Aguilar Vulcano Workshop 2010

IC40 MUON SAMPLE

L3 (muon sample) Trigger L2 (muon filter)

Strings Year Livetime SMT rate (Hz) µ filter rate (Hz)

  • atm. ν

final rate 9 2006 137 d 80 6 1.7 22 2007 276 d 450 20 18/d 40 2008 375.5 d 1100 23 40/d 59 2009 320 d 1900 24 79 2010 2300 40

5

IC40 rate evolution Detector rates:

  • Atm. neutrinos
  • Atm. muons

Get all sky cutting on total charge

IC40 All-Sky Stacking Conclusions Upgoing Downgoing

slide-6
SLIDE 6

Juan Antonio Aguilar Vulcano Workshop 2010

ANGULAR RESOLUTION

Cumulative distribution of the Point Spread Function for different IceCube configurations:

50% angular resolution

6

IC40 All-Sky Stacking Conclusions

Top view

  • IC40 has already in
  • ne direction a size

equals to IceCube (IC80) IC40

slide-7
SLIDE 7

Juan Antonio Aguilar Vulcano Workshop 2010

EFFECTIVE AREA

Solid-angle averaged neutrino effective area:

The smaller the zenith, the more sensitive to higher energies.

South Pole

180-150 150-120 120-90 60-90 30-60 0-30

IC40 All-Sky Stacking Conclusions There is a factor 2 increase at high energy and > 8 at low energies.

IC80+DC6, ϴ = (120◦ - 90◦) IC80+DC6, ϴ = (150◦ - 120◦) IC80+DC6, ϴ = (180◦ - 150◦) IC40, ϴ = (120◦ - 90◦) IC40, ϴ = (150◦ - 120◦) IC40, ϴ = (180◦ - 150◦)

slide-8
SLIDE 8

8

ALLSKY ANALYSIS

slide-9
SLIDE 9

Juan Antonio Aguilar Vulcano Workshop 2010

LIKELIHOOD AND DENSITY FUNCTIONS

Likelihood Space around track solution fit to paraboloid: width = σ

Signal pdf:

9

IC40 All-Sky Stacking Conclusions

!"#$ !" %&'()$

slide-10
SLIDE 10

Juan Antonio Aguilar Vulcano Workshop 2010

LIKELIHOOD AND DENSITY FUNCTIONS

Signal pdf:

10

IC40 All-Sky Stacking Conclusions

slide-11
SLIDE 11

Juan Antonio Aguilar Vulcano Workshop 2010

LIKELIHOOD AND DENSITY FUNCTIONS

Signal pdf:

Scrambled real data

Background pdf:

11

IC40 All-Sky Stacking Conclusions

slide-12
SLIDE 12

Juan Antonio Aguilar Vulcano Workshop 2010

LIKELIHOOD AND DENSITY FUNCTIONS

Likelihood: Background pdf: Signal pdf:

12

IC40 All-Sky Stacking Conclusions

slide-13
SLIDE 13

Juan Antonio Aguilar Vulcano Workshop 2010

LIKELIHOOD AND DENSITY FUNCTIONS

Likelihood: Background pdf: Maximization of the likelihood ratio:

logλ = log L(ˆ γ , ˆ n

s)

L(ns = 0)      

Estimates that maximize the Likelihood The final significance is determined by scrambling the data in r.a. and repeating the analysis. Signal pdf:

13

IC40 All-Sky Stacking Conclusions

slide-14
SLIDE 14

Juan Antonio Aguilar Vulcano Workshop 2010 14

Two main components in the final analysis sample:

  • The Southern sky (downgoing atm. muons) has a better angular resolution.
  • The Northern sky (upgoing atm. neutrinos) has a lower energy range.

Northern Sky Southern Sky

IC40 All-Sky Stacking Conclusions

IC40 ALLSKY ANALYSIS SAMPLE

slide-15
SLIDE 15

Juan Antonio Aguilar Vulcano Workshop 2010

Paper published by ApJL, 701:L47, 2009

24h +30° +15° +45° +60° +75°

  • log10 p
  • Atm. neutrinos

IC22 ALL SKY RESULTS

Hottest spot found at r.a. 153.4° , dec. 11.4° Post-trials p-value of analysis is ~ 1.34% (2.2 sigma)

15

IC40 All-Sky Stacking Conclusions

slide-16
SLIDE 16

Juan Antonio Aguilar Vulcano Workshop 2010

  • log10 p

IC22 ALL SKY RESULTS

0h 24h

+30° +15° +45° +60° +75° ‐15° ‐30° ‐45°

  • Atm. neutrinos
  • Atm. muons

Analysis extended to ‐50° dec

Paper published by Phys. Rev. Lett. 103, 221102 (2009)

First IceCube search for point sources above the horizon optimized for higher energies

16

IC40 All-Sky Stacking Conclusions

slide-17
SLIDE 17

Juan Antonio Aguilar Vulcano Workshop 2010

IC40 EVENT MAP

  • Atm. neutrinos
  • Atm. muons

Jon Dumm

17

IC40 All-Sky Stacking Conclusions

  • Livetime: 375.5 days
  • Events: 37290 (14139 up-going, 23151 down-going)
slide-18
SLIDE 18

Juan Antonio Aguilar Vulcano Workshop 2010

IC40 SIGNIFICANCE MAP

  • Atm. neutrinos
  • Atm. muons
  • Log-likelihood is calculated on a fine grid: 0.1° x 0.1°
  • Significance comes from the hottest single spot, calculated as the fraction
  • f scrambled trials with equal or higher significance – robust result

18

IC40 All-Sky Stacking Conclusions

Preliminary

Jon Dumm

slide-19
SLIDE 19

Juan Antonio Aguilar Vulcano Workshop 2010

IC40 SIGNIFICANCE MAP

  • Atm. neutrinos
  • Atm. muons
  • Log-likelihood is calculated on a fine grid: 0.1° x 0.1°
  • Significance comes from the hottest single spot, calculated as the fraction
  • f scrambled trials with equal or higher significance – robust result

19

IC40 All-Sky Stacking Conclusions

Preliminary

Jon Dumm

Ho#est loca+on in the all‐sky search is: Ra=115.05, Dec=15.05

slide-20
SLIDE 20

Juan Antonio Aguilar Vulcano Workshop 2010

IC40 ALLSKY RESULTS

  • The distribution of results for 10000

scrambled skymaps.

  • For each scrambled trial, the same

analysis is performed.

  • The hottest spot in each trial yields

the -log10(pmin) that goes into this

  • histogram. 9588 of the 10000

scrambled skymaps had a -log10(pmin) equal or greater than that of the real dataset ➜ all‐sky p‐value = 96%.

20 16

IC40 All-Sky Stacking Conclusions

  • bserved value

96%

slide-21
SLIDE 21

Juan Antonio Aguilar Vulcano Workshop 2010

UPPER LIMITS (90% CL)

21

  • Median Neyman

sensitivities/UL for E-2 neutrino spectrum

  • Discovery potential at

5σ (P = 50%) takes ~3x more flux

  • IceCube (IC86) ~2x

better than IC40

Preliminary

IC40 All-Sky Stacking Conclusions

slide-22
SLIDE 22

SPECTRAL INDEX

1-sigma

E-3 E-2 E-1.5

22

Number of events for a 50% discovery potential

How well we can fit the spectral index depends on the number of signal events. Softer spectrum require higher number

  • f events to claim a a

discovery.

IC40 All-Sky Stacking Conclusions

slide-23
SLIDE 23

Juan Antonio Aguilar Vulcano Workshop 2010

  • Analysis is optimized for E-2

spectra.

  • Gamma spectra from galactic

sources exhibit cut-offs at energies of about 5-30 TeV.

  • Southern sky is more sensitive

to a cut-off. But we are still sensitive in the Northern hemisphere even if we loose sensitivity for cut-offs < 100 TeV

23

IC40 All-Sky Stacking Conclusions

SPECTRUM CUTOFF

dΦ/dE = Φ0E −2 exp(−E /Ecutoff )

slide-24
SLIDE 24

24

STACKING ANALYSIS

slide-25
SLIDE 25

Juan Antonio Aguilar Vulcano Workshop 2010 25

STACKING ANALYSIS

  • Combine the signal from similar sources: Stacking
  • Small modification in the PS search:
  • The gain by adding more identical sources is

somewhere between:

  • √nSources ➜ expectation for high background
  • nSources

➜ expectation for low background

IC40 All-Sky Stacking Conclusions Sum of signal pdf

slide-26
SLIDE 26

Juan Antonio Aguilar Vulcano Workshop 2010

MILAGRO-FERMI SOURCES: THE GEV – TEV CONNECTION

Abdo et al. 2009, arXiv:0904.1018v3 26

IC40 All-Sky Stacking Conclusions

  • Milagro showed that many Fermi GeV gamma ray sources extend to TeV
  • Of 34 galactic Fermi BSL sources in Milagro 6 (14) are observed with > 5 (3) σ
  • Probability of 4 or more excesses > 3 σ in 34 trials is p = 1.5 x 10-7
  • Strong evidence for multi-TeV emission from Galactic BSL sources as a class.
slide-27
SLIDE 27

Juan Antonio Aguilar Vulcano Workshop 2010

NEARBY CLUSTER OF GALAXIES

27 Name Size (°)RA(°) Dec(°) Virgo 8.0 186.6 12.7 Centaurus 1.2 193.2 ‐41.3 Perseus 2.4 50.0 41.5 Coma 1.8 195.0 28.0 Ophiuchus 0.9 247 ‐24.5

IC40 All-Sky Stacking Conclusions

Murase et al. 2008, arXiv:0805.0104v2:

  • Cluster of Galaxies (CG)

might explain CR between the second knee and the ankle.

  • Different models CR

density predicts neutrino emission.

  • Detection of individual

neutrino signal is still challenging for IceCube-like detectors ➜ Stacking.

slide-28
SLIDE 28

Juan Antonio Aguilar Vulcano Workshop 2010

Stacked Sources (# of sources) P-value from scrambling Fit Number

  • f Source

Events Fit Spectral Index 90% Upper Limit

(10−12TeV−1cm−2s−1)

Milagro (17) 41.2% 4.2

  • 2.8

8.0 Nearby Clusters (5) 82.0% 0.6

  • 3.9

10.6 Starbursts Galaxies (127) 28.1% 2.9

  • 2.3

36.5

IC40 STACKING RESULTS

Dense ISM ➜ Elevated star formation rates ➜ SNR ➜ CR ➜ Neutrinos Becker et al. 2009, arXiv:0901.1775v1

28

IC40 All-Sky Stacking Conclusions

Preliminary

slide-29
SLIDE 29

Juan Antonio Aguilar Vulcano Workshop 2010

STACKING 6 MILAGRO SNR

IC40 Stacking Search

  • Med. Sensitivity

90% Upper Limit Milagro 6 SNR 2.05 * prediction 5.50 * prediction

3.0 events in IC40 predicted by flux from Halzen et al. 2008, arXiv:0803.0314v2

AMANDA 7-yr IC22 IC40 20% 27% 2.3%

p-values of 6 Milagro SNR stacked searches: (a posteriori)

29

IC40 All-Sky Stacking Conclusions

Preliminary

Will be tested again in future detector configurations…

slide-30
SLIDE 30

30

CONCLUSIONS

slide-31
SLIDE 31

Juan Antonio Aguilar Vulcano Workshop 2010

CONCLUSIONS

  • IceCube is starting the IC79 data taking and IC59 data analysis has started.
  • The whole year of IC40 has been analyzed for point-source and no

evidence of a neutrino point source has been found.

  • By stacking similar sources we can enhance the sensitivity/discovery

potential.

  • IceCube has already a sensitivity 10 times better than any neutrino

telescope in its preferred hemisphere and with IC40 we have started the extension to all-sky analysis.

31

IC40 All-Sky Stacking Conclusions