1. 1 Multi-TeV Observation on the Galactic Cosmic Ray Anisotropy in the Tail-In and Cygnus Regions by the Tibet-III Air Shower Array C. T. Yan 08 / 12 / 2006 Inst. for Cosmic Ray Research, Univ. of Tokyo ” Locating PeV Cosmic-Ray Accelerators: Future Detectors in Multi-TeV Gamma-Ray Astronomy ” 6 – 8 December, 2006 - Adelaide, Australia For the Tibet AS  Collaboration **

2. 2 The Tibet AS  Collaboration M.Amenomori, 1 S.Ayabe, 2 X.J.Bi, 3 D.Chen, 4 S.W.Cui, 5 Danzengluobu, 6 L.K.Ding, 3 X.H.Ding, 6 C.F.Feng, 7 Zhaoyang Feng, 3 Z.Y.Feng, 8 X.Y.Gao, 9 Q.X.Geng, 9 H.W.Guo, 6 H.H.He, 3 M.He, 7 K.Hibino, 10 N.Hotta, 11 HaibingHu, 6 H.B.Hu, 3 J.Huang, 12 Q.Huang, 8 H.Y.Jia, 8 F.Kajino, 13 K.Kasahara, 14 Y.Katayose, 4 C.Kato, 15 K.Kawata, 12 Labaciren, 6 G.M.Le, 16 A.F. Li, 7 J.Y.Li, 7 Y.-Q. Lou, 17 H.Lu, 3 S.L.Lu, 3 X.R.Meng, 6 K.Mizutani, 2,18 J.Mu, 9 K.Munakata, 15 A.Nagai, 19 H.Nanjo, 1 M.Nishizawa, 20 M.Ohnishi, 12 I.Ohta, 21 H.Onuma, 2 T.Ouchi, 10 S.Ozawa, 12 J.R.Ren, 3 T.Saito, 22 T.Y.Saito, 23 M.Sakata, 13 T.K.Sako, 12 T.Sasaki, 10 M.Shibata, 4 A.Shiomi, 12 T.Shirai, 10 H.Sugimoto, 24 M.Takita, 12 Y.H.Tan, 3 N.Tateyama, 10 S.Torii, 18 H.Tsuchiya, 25 S.Udo, 12 B. Wang, 9 H.Wang, 3 X.Wang, 12 Y.G.Wang, 7 H.R.Wu, 3 L.Xue, 7 Y.Yamamoto, 13 C.T.Yan, 12 X.C.Yang, 9 S.Yasue, 26 Z.H.Ye, 16 G.C.Yu, 8 A.F.Yuan, 6 T.Yuda, 10 H.M.Zhang, 3 J.L.Zhang, 3 N.J.Zhang, 7 X.Y.Zhang, 7 Y.Zhang, 3 Yi Zhang, 3 Zhaxisangzhu, 6 and X.X.Zhou 8 (1) Dep. of Phys., Hirosaki Univ., Hirosaki, Japan (2) Dep. of Phys., Saitama Univ., Saitama, Japan (3) Key Lab. of Particle Astrophys., IHEP, CAS, Beijing, China (4) Fac. of Eng., Yokohama National Univ., Yokohama, Japan (5) Dep. of Phys., Hebei Normal Univ., Shijiazhuang, China (6) Dep. of Math. and Phys., Tibet Univ., Lhasa, China (7) Dep. of Phys., Shandong Univ., Jinan, China (8) Inst. of Modern Phys., South West Jiaotong Univ., Chengdu, China (9) Dep. of Phys., Yunnan Univ., Kunming, China (10) Fac. of Eng., Kanagawa Univ, Yokohama, Japan (11) Fac. f of Educ., Utsunomiya Univ., Utsunomiya, Japan (12) ICRR., Univ. of Tokyo, Kashiwa, Japan (13) Dep of Phys., Konan Univ., Kobe, Japan (14) Fac. of Systems Eng., Shibaura Inst. of Tech., Saitama, Japan (15) Dep. of Phys., Shinshu Univ., Matsumoto, Japan (16) Center of Space Sci. and Application Research, CAS, Beijing, China (17) Phys. Dep. and Tsinghua Center for Astrophys., Tsinghua Univ., Beijing, China (18) Advanced Research Inst. for Sci. and Engin., Waseda Univ., Tokyo, Japan (19) Advanced Media Network Center, Utsunomiya University, Utsunomiya, Japan (20) National Inst. of Info., Tokyo, Japan (21) Tochigi Study Center, Univ. of the Air, Utsunomiya, Japan (22) Tokyo Metropolitan College of Industrial Tech., Tokyo, Japan (23) Max-Planck-Institut fuer Physik, Muenchen, Germany (24) Shonan Inst. of Tech., Fujisawa, Japan (25) RIKEN, Wako, Japan (26) School of General Educ.,Shinshu Univ., Matsumoto, Japan

3. 3 Outline The Tibet air-shower array Anisotropy of galactic cosmic rays (*)Anisotropy of galactic cosmic rays (*) The tail-in and loss-cone model Gamma/hadron separation methodGamma/hadron separation method –Discrimination of gamma/hadron in the array –Gamma/Hadrons judgment by comparisons (**) Back-check by the Crab Nebula Investigation on two anisotropy componentsInvestigation on two anisotropy components –The ‘tail-in’ anisotropy component –Excesses from the Cygnus region (***) Conclusive remarks

4. 4 The Tibet air shower array View around the Tibet III array (90.52E, 30.10N;4300m a.s.l.) in 2003 –Located at an elevation of 4300 m (Yangbajing in Tibet, China) –Atmospheric depth 606 g / cm 2 –Wide field of view ( ~ 2 sr field of view) –High duty cycle ( > 90%) –Modal energy: ~ 3 TeV –Angular resolution: ~ 0.9 o –Data sample used (1997 ~ 2005, 37 * 10 9 ) Large-scaleobservation

5. 5 Anisotropy of galactic cosmic rays 4.0 TeV 6.2 TeV 12 TeV 50 TeV 300 TeV From Science, V314, pp.439 – 443 (2006), by the analysis method (I) i) Temporal variation Anisotropy towards the Cygnus region ii) Anisotropy towards the Cygnus region iii) Energy dependency iv) Anisotropy fade away ~ 300 TeV

6. 6 loss-cone tail-in Galactic plane Tail-in and loss-cone model of the anisotropy 1)Heliospheric magnetic field is not enough for TeV CR anisotropy. 2)TeV CR anisotropy should be caused by the Local Interstellar Could (~ a few pc). R L ~ 0.01pc (for 10TeV proton in 1mG) Ref: K. Nagashima, K. Fujimoto, R.M. Jacklyn, J. Geophys. Res. V103, 17429 (1998). < 1 TeV

7. 7 The Gamma/Hadron Separation Gamma-initiated air shower –Concentrated –Smooth –Uniformity –… Hadron-initiated air shower –Scattered –Large fluctuation –Sub core structure –… Simulations: –Corsika-6.204 for air showers –Epicsuv-8.00 for array detectors –Energy: 300 GeV – 10 PeV –E -2.7, Hadrons (comp. HD4) –E -2.6, Gamma (Crab-like) Data cuts: –Zenith < 45 o –Core inside array –Residual error < 1.0 m –1.25 p / any 4 –30.0 < Sum_pFT <= 100.0 Representative energy:Representative energy: –4.2 TeV, Gamma –8.1 TeV, Hadrons Angular resolutionAngular resolution –0.9 o Data sample here (1999 ~ 2004, 10 * 10 9 )

8. 8 Discrimination parameter for gamma and hadrons R0 distributions & survival ratios Separation parameter –Global parameter Mean distance to core Virial distance of shower Hit_max to core Out core / All –Cluster parameter Num_clus / Num_hit Lateral distance of clus Steepness of clus Out_pixel / all_pixel –Image (FFT) parameter 1 st freq / DC 2 nd freq / DC 1 st freq / All 2 nd freq / All where R i is the distance between i th fitted detector and shower core in the shower-front plane. Gamma (MC) Hadron (MC) Real Data Syst ~= 5%

9. 9 Rejection method –Excess to Bkgrd Ratio (E2BR): E/B: E2BR before cut E’/B’: E2BR after cut Gamma survival ratio Hadron survival ratio –Expectation: 100% gamma: E = E’ 100% hadron: E = E’’ Where E’’ = ** E’ –Hypothesis Rejection: 100% gamma by Quality factor1 100% hadron by Quality factor2 Gamma/hadron rejection and its quality factors Quality factors: The key point is to compare the data sample before cut and after cut !!

10. 10 ~ 0.0046 +/- 0.00085 ~ 0.0077 +/- 0.0012 ~ 0.0031 +/- 0.00091 MC expected: 0.0069 +/- 0.00012 100% gamma-ray excess, Hadron (100%) is rejected at 3.4 sigma; Data is consistent with gamma(100%) at 0.8 sigma. Back-Check by the Crab Nebula Data analysis by azimuth swapping (The standard gamma-ray source) Before Cut After Cut Comparison (b) (a) (b) Bin size: 1.7deg * 1.7 deg

11. 11 Investigations on Two Anisotropy Components: the Tail-In and Cygnus regions 3.0 deg smoothed Before Cut After Cut Comparison 100% CR excess assumption100% gamma excess assumption Data analysis by weighted azimuth swapping

12. 12 Hints on the Tail-in and the Cygnus excesses Gamma-like Hadron-like Large-scale anisotropy removed b = -5 b = +5 b = -5 b = +5 Search Region (II) Search Region (I) Comparison Comparison 100% CRs Ex. 100% Gam. Ex. Tail-InCygnus

13. 13 Investigation on the Tail-In anisotropy component:Investigation on the Tail-In anisotropy component: Is it CR !? Use it as the background source ( Is it CR !? ) (Independent) bin size: 10 deg * 12 deg “Tail-In” Region Region

14. 14 ~ 0.0025 +/- 0.00014 ~ 0.0025 +/- 0.00028 ~ 0.0011 +/- 0.00015 Gamma (100%) is rejected at 7.4 sigma; Data is consistent with hadron (100%) at 0.1 sigma. If 100% gamma-ray, from the MC expectation. reduced Excess to Background Ratio = 0.0014 +/- 0.0008, Before Cut After Cut Comparison Is it CR !? ( Is it CR !? The GeV underground muon Exp. gives the answer is Yes !)

15. 15 Investigation on the excess from the Cygnus regionInvestigation on the excess from the Cygnus region ( gamma point source, diffuse gamma-ray emissions ) After g/p cut, Excess to Background Ratio will be enhanced, if excess is from gamma. See next  MGRO J2019+37 ( Cross + is MGRO J2019+37) +/- 3.0 deg Before Cut After Cut On Off1 Off2 b = +5 b = -5 3.0 deg smoothed

16. 16 Hadron (100%) is rejected at 3.4 sigma; Data is consistent with gamma (100%) at 1.8 sigma. Hadron Rejection:on & off Hadron Rejection: on & off the Galactic plane on Cygnus Region ~ 0.00065 +/- 0.00022 ~ 0.00198 +/- 0.00045 ~ 0.00133 +/- 0.00040 MC expected: 0.00120 +/- 0.00045 If 100% gamma, off1off2 Comparison After Cut Before Cut

17. 17 Conclusive Remarks The Crab excess is consistent with 100% gamma assumption at 0.8 sigma level, and 100% CRs assumption is rejected at 3.4 sigma level. (The CRs rejection is MC-independent). The Tail-In region anisotropy is from CRs except the small region including the Crab Nebula. 100% gamma excess assumption is rejected at 7.4 (3.5 [large-scale anisotropy removed]) sigma level. And 100% CR excess assumption is consistent at 0.1 (0.4 [large- scale anisotropy removed]) sigma level. As the original excess from the Cygnus region in our search window (-4.0 o < b < 2.0 o, 72.0 o < l < 78.0 o ) is at 3.3 sigma level, we cannot effectively judge it is from gamma-ray or CRs. CRs rejection is at about 3.4 sigma level. And the gamma-ray consistence is at about 1.8 sigma level. Due to the fluctuations, here the result shows the excess is over gamma-like. But the (diffuse) gamma-ray emission hypothesis is slightly favored. Further improvement using multi-parameters is in progress.

18. 18 Appendix: background estimations Global CR intensity fitting methods (I), (II) Technique of time swapping (from Milagro) Azimuth swapping method Weighted azimuth swapping method

19. 19 Global CR intensity fitting method (I) Reference: M. Amenomori, et al., ApJ. V633, 1005 (2005) Used in the published result

20. 20 Global CR intensity fitting method (II) The background is estimated by weighted azimuth swapping A Technique of Data Shuffling Auto event and background normalization 1) Auto event and background normalization Auto azimuth correction in swapping M.C. 2) Auto azimuth correction in swapping M.C.