GeV GeV GeV MeV 300GeV GeV 50GeV 20-50% TeV 2020 Cherenkov Telescope Array (CTA) 20 GeV 0 TeV CTA 1 TeV 23 m 1855 ( ) 2000 GeV GeV

Transcription

1 CTA GeV 0GeV

2 GeV GeV GeV MeV 300GeV GeV 50GeV 20-50% TeV 2020 Cherenkov Telescope Array (CTA) 20 GeV 0 TeV CTA 1 TeV 23 m 1855 ( ) 2000 GeV GeV

3 GeV GeV GeV (TKR) (CAL) (ACD) ACD Pass

4 Mrk Cherenkov Telescope Array PMT (Q.E.) F CTA Single p.e Multi p.e F F

5 5.6.2 PACTA vs F vs F

6 V. Hess X TeV 20 ev ev 16 ev Knee; 18.5 ev Ankle; IC 443 W44 - π 0 4

7 4 Knee Grigorov 2nd Knee sr -1 ] 1.6 s -1 m -2 F(E) [GeV 2.6 E 3 2 JACEE MGU Tien-Shan Tibet07 Akeno CASA-MIA HEGRA Fly s Eye Kascade Kascade Grande IceTop-73 HiRes 1 HiRes 2 Telescope Array Auger Ankle E [ev] : [1] E 2.6 5

8 [2] 15 ev 8 16 ev (Active Galactic Nuclei: AGN) (Gamma-Ray Burst: GRB) GeV X GeV 1991 Energetic Gamma Ray Experiment Telescope (EGRET) kev - GeV MeV GeV EGRET 6

9 EGRET NaI (Anti-Coincidence Detector: ADC) 2008 GeV (Large Area Telescope: LAT) EGRET 1.2 LAT 0 MeV GeV The Astrophysical Journal Supplement Series, 199:31(46pp),2012April Nolan et al Figure 1. Sky map of the energy flux derived from the LAT data for the time range analyzed in this paper, Aitoff projection in Galactic coordinates. The image shows γ -ray energy flux for energies between 0 MeV and GeV, in units of 7 erg cm 2 s 1 sr : 0 MeV - GeV [3] Hammer-Aitoff These Pass 7_V6 (P7_V6) Source 90 class event selections are accompanied by a corresponding revised set of IRFs (A. A. Abdo et al. 2012b, in preparation), including an energy-dependent PSF calibrated using known celestial point sources. The model for the diffuse gamma-ray background was fit using P7_V6 Clean event selections and IRFs (see Section 2.2). The Clean event selection has lower residual background intensity than P7_V6 Source at the cost of decreased effective area, a tradeoff that is worthwhile for studies of diffuse γ -ray emission. The IRFs tabulate the effective area, PSF, and energy dispersions as functions of energy and inclination angle with respect to the LAT z-axis. The IRFs are also tabulated as a function of the location of the γ -ray conversion in the LAT; Front conversions occur in the top 12 tracking layers. The tungsten foils are thinnest in this region and the PSF is narrower than for the Back section, which has four layers of relatively thick conversion foils. The 2FGL catalog is therefore derived from a new data set rather than simply an extension of the 1FGL data set. During the 1FGL time interval (up to 2009 July 4) the standard rocking angle for survey-mode observations was 35. During much of 2009 July and August it was set to 39.Thenon2009 September 2 the standard rocking angle was increased to 50 in order to lower the temperature of the spacecraft batteries and thus extend their lifetime. Time intervals during which the rocking angle of the LAT was greater than 52 were excluded. The more conservative 1FGL limit of 43 had to be raised to accommodate the larger standard rocking angle. For the 2FGL analysis we apply a more conservative cut on the zenith angles of the γ -rays, 0 instead of the GeV with decreasing energy, a more conservative limit on zenith angle is warranted in any case. The energy flux map of Figure 1 summarizes the data set used for this analysis. The corresponding exposure is relatively uniform, owing to the large field of view and the rockingscanning pattern of the sky survey. With the new rocking angle set to 50 the exposure is minimum at the celestial equator, maximum at the north celestial pole and the contrast (maximum to minimum exposure ratio) is 1.75 (Figure 2). The exposure with rocking angle 35 (Figure 2 of Abdo et al. 2009d) was least at the south celestial pole, with a contrast of The north/south asymmetry is due to loss of exposure during passages of Fermi through the SAA. Figure 3 shows that the original rocking scheme resulted in a very uniform exposure over the sky. The new rocking scheme is less uniform, although it still covers the entire sky to an adequate depth. The exposure map for 2FGL is about halfway between the 35 and 50 maps. It peaks toward the north celestial pole and is rather uniform over the south celestial hemisphere, with a contrast of Note that the average etendue of the telescope is only slightly reduced, from 1.51 m 2 sr (at 1 GeV) in the first 11 months to 1.43 m 2 sr over the last 11 months. The reduction is due to the part of the field of view rejected by the newer zenith angle selection. TeV Model for the Diffuse Gamma-Ray Background The γ -ray emission produced by the Galaxy originating from the interaction of cosmic-ray electrons and protons with interstellar nucleons and photons is modeled with the same

10 1TeV 50 /m nm Photomultiplier Tube: PMT ( 1.3 ) ( 1.3 ) 16 Very high energy gamma-ray detectors : [4] 1.5 Hillas 1.5 Hillas LENGTH WIDTH DIST ALPHA - 8 Figure 2.1. Main panel: Monte Carlo simulations of 320 GeV gamma-ray shower and

11 1.4: 0GeV [5] 9

12 50-0 p.e. 4. Observation Technique and Data Analysis 67 [7] ALPHA center of field of view DIST LENGTH WIDTH COG Figure 4.18: Definition 1.5: ofmagic image parameters, defined for ahillas moment [6] analysis of a shower image in the camera plane Event Classification with Random Forest and Estimation ofthe Energy 1989 Whipple m 9σ TeV The great bulk of events triggering the telescope comes from cosmic hadrons. Efficient cuts are therefore needed to separate γ-rays and hadrons and thus to reduce the background. Usually, these cuts are based on the image (and time) parameters, because these parameters are rather robust under changing observation 5conditions - km (e.g. different NSB levels, weather changes, clouds) In order to 1.6 obtain optimal results, these cuts have to be determined in a high dimensional parameter space. In the standard 1.7 analysis, thistaskisaccomplishedby the RF algorithm, based on multidimensional decision trees. TheRFistrainedwithMC γ-rays as signal events. As background, a measured 40% data sample 15% fromaposition[25] pointing to a region in the sky where no γ-rays are expected 0 m (so-called OFF data sample) is used. Each event (background and signal event) is described by a set of(image) parameters forming a vector. A RF consists of a number of independent binary decision trees based on those parameters. The algorithm performs the following steps: 1. A sample of both, signal and background data, is drawn. 2. A random (image) parameter is chosen. The cut in this parameter that minimizes the Gini index 14 is applied and thus the data sample is split into two sets, forming two nodes. 3. For each of the two subsets, another random parameter is chosen and the cut applied

13 3.1.4 u b 1.6: [7] z cu b m ( )z c 91 11

14 1.7: 5- km 12

15 H.E.S.S. MAGIC VERITAS 1.8 MAGIC TeV : MAGIC 1.9: TeV [9] GeV GeV 1. 13

16 1.: [8] COMPTEL EGRET 1991 Fermi MAGIC H.E.S.S. CTA

17 19 0 LAT EGRET 1500cm 2-0GeV 8000cm 2 ( ) 300 GeV [] LAT 1FHL [3] LAT GeV GeV H.E.S.S. MAGIC VERITAS m MAGIC 17m 50 GeV [11] Cherenkov Telescope Array (CTA) GeV Counts per (0.5 deg) 2 Fig. 1. Sky map of γ-ray counts above GeV in Galactic coordinates in Hammer-Aitoff projection. The Galactic center (0,0) is at the center of the map and Galactic longitude increases to the left. The binning is 0. 5andtheimagehasbeensmoothedwitha2-dimensionalGaussian of full width at half maximum : GeV [3] Hammer-Aitoff 2.4. Exposure, Diffuse Gamma-Ray Backgrounds, and Point-source Sensitivity The time interval analyzed here is from the beginning of science operations, 2008 August 4 (MET ) to 2011 August 1 (MET ), covering very nearly 3 years 4.Theoverall exposure for the 3-year interval is relatively uniform (Fig. 2), rangingfrom 15% to +38% of the average value of 9.5 cm 2 s, primarily as a function of declination. The exposure at southern declinations is somewhat less because no observations are made during passages through the South Atlantic Anomaly. In addition, the exposure near the northern celestial pole is enhanced because the majority of non-survey mode (pointed) observations have beenmadetowardnortherntargets. The exposure is slightly depressed in a 21 diameter region near the southern celestial pole because of the 5 limit on zenith angle for the γ rays selected for analysis ( 2.3) GeV (Active Galactic Nuclei: AGN) GeV 15 Proper quantification of the diffuse backgrounds is necessary for accurate source detection and characterization. We used the publicly available models forthegalacticandisotropicdiffuse emissions for this analysis. These files, gal 2yearp7v6 v0.fits and iso p7v6clean.txt, canbe 4 Mission Elapsed Time (MET), the number of seconds since 00:00UTCon2001January1(excludingleapseconds).

18 AGN AGN 6 9 AGN AGN AGN 1.12 AGN 1.12: AGN [12] AGN AGN AGN Markarian 421 (Mrk 421)

19 π : Mrk421 [14] 2009/1/ /6/1 GeV-TeV X 2 LAT GeV [13] GeV GeV z 2 AGN (Extra-galactic Background Light: EBL) 1.14 z=1 0 GeV 1/e TeV 0 GeV 17

20 AGN EBL 1.14: EBL [15] τ τ = 1 1/e 11 z=0.003, 0.01, 0.03, 0.1, 0.3, 0.5, 1, 1.5, 2, 2.5, 3, (WIMP) WIMP [16] 1.15 WIMP GeV TeV [17] 18

21 1.15: DM [18] WIMP WIMP [16] LAT 2012 C. Weniger 130 GeV 3.2σ [19] LAT [20] LAT (Gamma-Ray Burst: GRB) MeV [21] 19

22 44 J (= 51 erg) GRB GRB GRB GRB?? 1.16: GRB X [22] LAT MeV GeV 40 GRB [23] GeV [24] GeV 20

23 for the highest-energy γ-rays associated with LAT GRBs. The estimated errors are computed from the energy dispersion in the Instrument Response Functions and it is of the order of % for energies >1GeV. When possible, we also indicate the source frame energy. Energy [GeV] ArrivalTime [s] Fig. 12. Observed (upward triangles) and ties as functions of rest-frame 1.17: GRB rest frame (downward triangles) energy and GRB as the best fitting broken po GRB arrival time for highest-energy events associated with long (blue) and short (red) LAT are very important to constra els. The later points in the [23] detected GRBs. Vertical dashed lines connect but they also would be the mo the observed and the rest frame energy for the any unaccounted-for systematic 1.4 same burst. Data points are from Table 8. arising, for example, from th estimation or the exposure ca the bottom panel of Fig. 13 w Temporally Extended Emission GeV - GeV the luminosity as a function of time, but for all the GRBs in t To study the temporal decay of the extended emission detected by the LAT, we sis. For the three GRBs with te In Table 9 we report the results utilized the time-resolved analysis described we report the decay index sta in 3.5. We first visualized any detected CTA peak flux and before the break extended emission using flux light curves LAT index after the break α 2, and t (shown in Appendix B), and then calculated CTA t b. For all other GRBs, we rep the peak-flux value F p CTA and the time of the index for the whole extended e peak flux t p, quantities shown in the two top ing from the peak flux, and th panels of Fig. 14. In the time-resolved analysis we adaptively changed the size of the time the low-energy (GBM) emissio for the light curve starting fro LAT CTA bin width in order to significantly detect the LAT Referring to Table 9, we a source, so F p corresponds tolat the average flux late-time decay index α L, CTA 37 CTA LAT 3 4 GRBs. The 4 most luminous burst the LAT have some of the high in the ensemble, all exceeding 1 Among the rest of the bursts, G and 1721 also have notab fluxes. GRB 0728A was at t FoV at the time of the GBM tr detected only at later times. It lowest peak flux of all GRBs, at of magnitude lower than the res lation; however, its value is po by large systematic errors. We also applied the method 3.5 to the subsample of G tected extended emission. We poral breaks in the decay of emission of three bright GRBs: GRB B and GRB panel of Fig. 13 we show t 21

24 (Large Area Telescope: LAT) (Gamma-ray Burst Monitor: GBM) LAT 20 MeV 300 GeV 1/4 2.1 LAT 2.1: LAT 20 MeV GeV 7000cm 2 (1GeV) 8000cm 2 (0GeV) 1.7m 2 sr (1GeV) 1.9m 2 sr (0GeV) 2sr 0.8 (1GeV) 0.2 (0GeV) < % (1-0 GeV) Pass6 (P6) Pass7 (P7) Pass8 (P8) 2.2 (Extra-galactic Gamma-ray Background: EGB) 2.1 SOURCE Pass7 SOURCE

25 2.2: LAT P7TRANSIENT P7SOURCE P7CLEAN P7ULTRACLEAN : LAT Pass7 SOURCE

26 2.2: LAT Pass7 SOURCE 4 0 MeV 2 phcm 2 s LAT 1.2 (Tracker: TKR) (Calorimeter: CAL) (Anti-Coincidence Detector: ACD) 2.2 TKR CAL 4x4 ACD (TKR) LAT (TKR) EGRET SLAC 0µm [25] 3% 12 TKR TKR 24

27 The Large Area Telescope on the Fermi Gamma-ray Space Telescope Mission IG. 1. SchematicdiagramoftheLargeAreaTelescope.Thetelescope s dimensions are 1.8 m 1.8 m 0.72 m. The power required and the mas 0 W and 2,789 kg, respectively. 2.3: LAT 3 [] TKR CAL 16 ADC TKR front section 12 3% X 0 TKR back section 4 18% X 0 2 no W CAL 2.4: LAT [26] X 0 25

28 X Y l l l l l l F IG. 5. Illustration of tracker design principles. The first two points dominate the measurement of the photon direction, especially at low energy. (Note that in this projection only the x hits can be displayed.) (a) Ideal conversion in W: Si detectors are located as close as possible to the W foils, to minimize the lever arm for multiple scattering. Therefore, scattering in the 2nd W layer has very little impact on the measurement. (b) Fine detector segmentation can separately detect the two particles in many cases, enhancing both the PSF and the background rejection. (c) Converter foils cover only the active area of the Si, to minimize conversions for which a close-by measurement is not possible. (d) A missed hit in the 1st or 2nd layer can degrade the PSF by up to a factor of two, so it is important to have such inefficiencies well localized and identifiable, rather than spread across the active area. (e) A conversion in the structural material or Si can give long lever arms for multiple scattering, so such material is minimized. Good 2-hit resolution can help identify such conversions. F IG. 6. LAT calorimeter module. The 96 CsI(Tl) scintillator crystal detector elements are arranged in 8 layers, with the orientation of the crystals in adjacent layers rotated by 90. The total calorimeter depth (at normal incidence) is 8.6 radiation lengths. 図 2.5: LAT の CAL[] カロリーメータ (CAL) 0.6 log(posadc/negadc) カロリーメータ (CAL)0.4には CsI 結晶シンチレータが採用されている シンチレータは mm 26.7 mm 19.9 mm の角材状になっており これを 12 本並べたものを１層 0 にする それを図 2.5 角材の方向を互い違いに８層重ねたもので各モジュールの CAL が 出来ている 図 角材の両端にはシンチレーション光を検出するフォトダイオードが高エネルギー用小さ -0.4 いものと低エネルギー用大きいものの２個ずつ取り付けられている どの角材で光ったか -0.6 と さらに両端の光量の比から電子または陽電子が通過した場所が分かり それが互い違 longitudinal position, mm いになっているので３次元的なシャワーの形が再構成出来るようになっている このシャ F IG. 7. Light asymmetry measured in a typical calorimeter crystal using sea level muons. The light asymmetry is defined as the logarithm of the ratio of the ACD とともにバックグラウンドを排除するのに用いられ outputs ワーの形状を再構成する能力は of the diodes at opposite ends of the crystal. The width of the distribution at each position is attributable to the light collection statistics at each end of the crystal for the 11 MeV energy depositions of vertically incident muons used in the analysis. This width scales with energy deposition as E 1/2. 次の章で述べる研究において最も重要な要素になる CAL の厚みは 8.6 放射長であり 高いエネルギーのイベントの場合シャワーの一部は CAL の後ろに抜けていく その割合は 0 GeV のイベントでは平均して半分ほどにもな る 数 GeV 以上のエネルギーではそうしたエネルギーの漏れがエネルギー分解能を制限 する支配的な原因になってしまう 反同時計数検出器 (ACD) 反同時計数検出器 (ACD) はタイル状もしくはリボン状のプラスチックシンチレータで ある 図 2.6 のように タイルが上面に 25 枚 側面には 16 枚ずつ配置されている リボ ンはタイルの隙間を埋めるためにタイルの下に張られている 側面の ACD が覆っている のは TKR までで CAL の側面はカバーしていない 2.3 解析の流れ LAT のデータは大まかに言って図 2.7 のようなが流れに沿って処理される 検出器の出 力結果がそのまま含まれる生データが物理的に意味のある高水準なデータに処理されてい 26

29 2.6: LAT ACD [] LAT CAL TKR ACD 2.7 Event classification ACD ACD ACD ACD 2.8 ACD CAL TKR ACD ACD 27

30 2.7: LAT CPF Charged particles in FoV 2.8: [30] ACD 28

31 2.4.2 ACD TKR ACD LAT LAT (Multi-Variate Analysis: MVA) Boosted Decision Tree (BDT) LAT Decision Tree ( ) 2.9 (depth) (training) [28] BDT Decision Tree [28] 29

32 8.12 Boosted Decision and Regression Trees 9 Figure 18: Schematic view of a decision tree. Starting from the root node, a sequence of binary splits using the discriminating variables x i is applied to the data. Each split uses the variable that at this node gives the 2.9: Decision Tree [28] best separation between signal and background when being cut on. The same variable may thus be used at several nodes, while others might not be used at all. The leaf nodes at the bottom end of the tree are labeled S for signal and B for background depending on the majority of events that end up in the respective nodes. For regression trees, the node splitting is performed on the variable that gives the maximum decrease in the average squared error when attributing a constant value of the target variable as output of the node, given by the average of the training events in the corresponding (leaf) node (see Sec ). 2.5 Pass8 Pass Booking options CAL The boosted decision (regression) treee (BDT) classifier is booked via the command: factory->bookmethod( Types::kBDT, "BDT", "<options>" ); CAL 1 TeV CAL 1 3TeV Code Example 50: Booking of the BDT classifier: the first argument is a predefined enumerator, the second argument is a user-defined string identifier, and the third argument is the configuration options string. Individual options are separated by a :. See Sec for more information on the booking. ACD CAL TKR ACD Several configuration options are available to customize the BDT classifier. They are summarized ACD in Option Tables 22 and 24 and described in more detail in Sec ACD 0σ ACD 5σ CAL CAL 30

33 4 4 th Fermi Symposium : Monterey, CA : 28 Oct-2 Nov 201 sr] 2 Acceptance [m Preliminary Event selection 1.25 P8SOURCE prototype P7SOURCE_V6 As the event reconstruction is nearing completion, the LAT Collaboration is now focusing on the next step of the event-level analysis, namely that responsible for the final determination of the high-level event properties, such as particle type, energy and direction. Similarly to what we did for the previous passes, we use Classification Trees (CTs) to select candidate gamma-rays on the basis of the reconstruction outputs. The particle identification CTs are trained using variables from all the three LAT subsystems. One noticeable difference is the use of the TMVA multivariate analysis framework [Hoecker et al. 2007]. Compared to the technology used in the current event classification, TMVA is capable of handling much larger data sets and allows for an overall faster development cycle. The CT performance is evaluated from the combination of background rate and gamma-ray acceptance that can be achieved for a given cut on the output signal probability. We note here that a differential background rate equal or 31 slightly lower than the Extragalactic Gamma-ray Background (EGB) rate is desirable for point-source analysis. We have studied several candidate Pass-8 event classes defined by event selections that allow varywe find a 25% increas Pass-7 source event clas 300 MeV) the increas as a factor three. 4. Extended Event In the current phot 0.5 track in the tracker or the calorimeter are sim of them are used in n Energy [MeV] the LAT Low-Energy Figure 1: Gamma-ray acceptance versus photon energy Pelassa et al. [20]. O 2.: for Pass8 the Pass-7 source class Pass7 and [29] a candidate Pass-8 event which is being investig class. is the development of tracker-only and calori these events have worse information in the background rejection removes outof-time signals from the ACD2. and provides a signifi- they can provide a ver with respect to those in cant increases in effective area. tive (Point area in some region Spread Function: PSF) might be exploited in s 4.1. Tracker-only E Below 0 MeV ma in the tracker and depo ter. Since seeing calor lated with tracks is a p reconstructed events a backgrounds that would tify, in the current ev minimum (5 MeV) ener However, the success o transients has made it carry useful informatio analysis. As the entire event se assessed in the context provide the potential fo effective area below 0 LAT phase space whic many science analyses Calorimeter-o While almost a half o have no usable tracke

34 3 3.1 (Large Area Telescope: LAT) (Tracker: TKR) (Calorimeter: CAL) (Anti-Coincidence Detector: ACD) TKR CAL 3.1 TKR 3.1 CAL TKR GeV [3] 3.1: CAL 32

35 TRANSIENT SOURCE CLEAN ULTRACLEAN CLEAN ULTRACLEAN SOURCE TRANSIENT (Extra-galactic Gamma-ray Background: EGB) EGB SOURCE EGB TRANSIENT EGB Pass7 LAT SOURCE 3 GeV 30 GeV 2.1m 2 sr 300 GeV 1.6m 2 sr Pass8 300 GeV 2.5m 2 sr 1 GeV TRANSIENT 40% - 50% 2.5m 2 sr 1.0m 2 sr Event Classification & Background Rejection Merit 1 33

36 3.2.1 CERN ROOT Toolkit for Multivariate Data Analysis with ROOT (TMVA) [28] TMVA LAT M. Wood Python TMVA 3.3 TKR CAL CAL CAL 30 GeV ACD : CalTrackAngle > 0.3!(TkrNumTracks > 0) Cal1RawEnergySum > Cal1MomNumIterations > 0 Cal1MomZDir > 0.2 TKR 0 TKR CAL 72.5 CAL 30 GeV CAL CAL

37 MC CalOnly event data sets Leptons Hadrons Gammas log(combined energy and position deviation in ACD) 3.2: Acd2Cal1VetoSigmaHit Acd2Cal1VetoSigmaHit Boosted Decision Tree (BDT) Acd2VetoSigmaHit ACD ACD Acd2VetoSigmaHit CAL ACD CAL ACD ACD CAL 0σ 5σ ACD 3.2 Acd2VetoSigmaHit Acd2VetoSigmaHit=0-4 35

38 MC CalOnly event data sets Leptons Hadrons Gammas Shower transverse size 3.3: Cal1TransRms Cal1TransRms ACD CAL Cal1TransRms 3.3 CAL ACD CalELayerCorrInitialRatio CalELayerCorrInitialRatio CAL CAL CalELayerCorrInitialRatio TMVA 36

39 3.2: CAL 0 CalTrSizeCalT95 Cal1TransRms CalEdgeEnergy Acd2TileEnergyRatio CalBkHalfRatio CalNewCfpCalSelChiSq Acd2VetoCount Acd2Cal1VetoSigmaHit Cal1MomNumCoreXtalsFract CalELayerCorrInitialRatio CalELayer34afterInitialRatio CalELayer74Ratio CalNewCfpCalTmax CAL RMS CAL CAL ACD ACD CAL CAL 0 mm χ 2 ACD ACD CAL CAL CAL CAL 3 4 CAL 4 Acd2Cal1Energy15 CAL 15 ACD Cal1FitChiSquare CAL χ 2 37

40 0.022 MC CalOnly event data sets Leptons Hadrons Gammas Magnitude of shower evolution 3.4: CalELayerCorrInitialRatio Acd2VetoSigmaHit 0 Acd2VetoSigmaHit > : < Cal1TransRms < 70 Acd2Cal1VetoSigmaHit > 0 ACD BDT TMVA 38

41 Gamma efficiency MC Hadron Zenith angle >= 53.1 deg Zenith angle < 53.1 deg MC Gamma Zenith angle >= 53.1 deg Zenith angle < 53.1 deg Background rejection MC Lepton Zenith angle >= 53.1 deg Zenith angle < 53.1 deg Cut value on the background classifier output 3.5: TMVA TMVA 3.5 TMVA 39

42 GeV GeV E -2 x Differntial Rate [MeV -1 sr s -1] CalOnly EGB Regular E -2-1 x Differntial Rate [MeV sr s -1] CalOnly EGB Regular Acceptance [m 2 sr] Acceptance [m 2 sr] E -2 x Differntial Rate [MeV -1 sr s -1] CalOnly EGB Regular GeV E -2-1 x Differntial Rate [MeV sr s -1] CalOnly EGB Regular GeV Acceptance [m 2 sr] Acceptance [m 2 sr] 3.6: 4 [m 2 sr] [MeVsr 1 s 1 ] 40

43 vs. TMVA GeV 316 GeV (Extra-galactic Gamma-ray Background: EGB) EGB 3.6 EGB EGB GeV 0.4m 2 sr m 2 sr 0.25m 2 sr % 1-20 % % - 13% 3%

44 0.06 Angular resolution MC energy and inclination angle bin GeV and θ < 53 (PSF68(95)=3.2 (6.6 )) 0-316GeV and θ < 53 (PSF68(95)=4.2 (9.8 )) GeV and θ > 53 (PSF68(95)=1.8 (3.5 )) GeV and θ > 53 (PSF68(95)=1.6 (3.8 )) Error in the reconstructed direction [deg] 3.7: Energy resolution MC energy and inclination angle bin GeV and θ < 53 ( E/E =%) 0-316GeV and θ < 53 ( E/E =14%) GeV and θ > 53 ( E/E =3.5%) GeV and θ > 53 ( E/E =3.3%) (Reconstructed energy - MC energy) / MC energy 3.8: 42

45 Number of remaining background components (MC) [events] 3 MC Source CrProtonPrimary 2 CrElectronPrimary CrPositronPrimary CrAlpha CrHeavyIon Cut value on the background classifier output 3.9: EGB 1 2 CalOnly1xEGB CalOnly2xEGB CalOnlyxEGB π 0 [26] 2 LAT 3. CalOnly1xEGB CalOnly2xEGB CalOnlyxEGB TRANSIENT SOURCE S/B / (S/B) S/B 3.4 ON/OFF ON OFF 43

46 [events] GeV Earthlimb P8R1_TRANSIENT_R0 P8R1_SOURCE CalOnly_xEGB CalOnly_2xEGB CalOnly_1xEGB [events] GeV Earthlimb P8R1_TRANSIENT_R0 P8R1_SOURCE CalOnly_xEGB CalOnly_2xEGB CalOnly_1xEGB Zenith theta [deg] Zenith theta [deg] [events] GeV Earthlimb P8R1_TRANSIENT_R0 P8R1_SOURCE CalOnly_xEGB CalOnly_2xEGB CalOnly_1xEGB [events] GeV Earthlimb P8R1_TRANSIENT_R0 P8R1_SOURCE CalOnly_xEGB CalOnly_2xEGB CalOnly_1xEGB Zenith theta [deg] Zenith theta [deg] 3.: 44

47 N ON N OF F A ON A OF F N S S/B N S = N ON N OF F A ON A OF F (3.1) S/B = N ON N OF F AON A OF F N OF F AON A OF F (3.2) GeV : ON/OFF ON ( ) OFF ( ) S/B TRNSIENT SOURCRE CalOnly1xEGB ON Mrk421 Mrk421 Mrk Mrk421 θ ON/OFF GeV CalOnly1xEGB CalOnly2xEGB SOURCE 24% S/B S/B

48 [events] Number of signal events Class P8R1_TRANSIENT_R0 P8R1_SOURCE CalOnly_xEGB CalOnly_2xEGB CalOnly_1xEGB log (Energy [MeV]) 3.11: 3.5: Mrk421 ON/OFF Mrk421 ON ( ) OFF ( )

49 S/B ratio S/B Class P8R1_TRANSIENT_R0 P8R1_SOURCE CalOnly_xEGB CalOnly_2xEGB CalOnly_1xEGB log (Energy [MeV]) 3.12: S/B CalTkrFilter P8R1_TRANSIENT_R0 P8R1_SOURCE DEC[deg] hmap0_0_0 Entries Mean x Mean y DEC[deg] hmap0_0_1 Entries 2145 Mean x Mean y DEC[deg] hmap0_0_2 Entries 380 Mean x Mean y RMS x RMS y RMS x RMS y RMS x RMS y GeV<=Energy<0.0GeV RA [deg] RA [deg] RA [deg] DEC[deg] CalOnlyFilter hmap0_1_0 Entries Mean x Mean y RMS x 5.77 RMS y DEC[deg] CalOnly_xEGB hmap0_1_1 Entries 869 Mean x 166 Mean y RMS x RMS y DEC[deg] CalOnly_2xEGB hmap0_1_2 Entries 151 Mean x Mean y RMS x RMS y DEC[deg] CalOnly_1xEGB hmap0_1_3 Entries 135 Mean x Mean y RMS x RMS y RA [deg] RA [deg] RA [deg] RA [deg] 3.13: Mrk GeV 4 ON 47

50 RA [deg] CalTkrFilter P8R1_TRANSIENT_R0 P8R1_SOURCE DEC[deg] hmap1_0_0 Entries Mean x Mean y DEC[deg] hmap1_0_1 Entries 2335 Mean x Mean y DEC[deg] hmap1_0_2 Entries 196 Mean x Mean y RMS x RMS y RMS x RMS y RMS x RMS y GeV<=Energy<177.8GeV RA [deg] RA [deg] DEC[deg] CalOnlyFilter hmap1_1_0 Entries Mean x Mean y DEC[deg] CalOnly_xEGB hmap1_1_1 Entries 742 Mean x Mean y DEC[deg] CalOnly_2xEGB hmap1_1_2 Entries 271 Mean x Mean y DEC[deg] CalOnly_1xEGB hmap1_1_3 Entries 148 Mean x 166 Mean y RMS x RMS y RMS x RMS y RMS x RMS y RMS x 4.29 RMS y RA [deg] RA [deg] RA [deg] RA [deg] 3.14: Mrk GeV 4 ON CalTkrFilter P8R1_TRANSIENT_R0 P8R1_SOURCE DEC[deg] hmap2_0_0 Entries Mean x Mean y DEC[deg] hmap2_0_1 Entries 1632 Mean x Mean y DEC[deg] hmap2_0_2 Entries 76 Mean x Mean y RMS x RMS y RMS x RMS y RMS x 3.26 RMS y GeV<=Energy<316.2GeV RA [deg] RA [deg] RA [deg] DEC[deg] CalOnlyFilter hmap2_1_0 Entries Mean x Mean y DEC[deg] CalOnly_xEGB hmap2_1_1 Entries 293 Mean x Mean y 37.5 DEC[deg] CalOnly_2xEGB hmap2_1_2 Entries 136 Mean x Mean y DEC[deg] CalOnly_1xEGB hmap2_1_3 Entries 116 Mean x Mean y RMS x RMS y RMS x RMS y RMS x RMS y RMS x RMS y RA [deg] RA [deg] RA [deg] RA [deg] 3.15: Mrk GeV 4 ON 48

51 GeV Class CalOnly_xEGB CalOnly_2xEGB CalOnly_1xEGB htheta0_1_1 htheta0_1_2 htheta0_1_3 Entries Mean RMS GeV Class CalOnly_xEGB CalOnly_2xEGB CalOnly_1xEGB htheta1_1_1 htheta1_1_2 htheta1_1_3 Entries Mean RMS Theta angle^2 [deg^2] Theta angle^2 [deg^2] GeV Class CalOnly_xEGB CalOnly_2xEGB CalOnly_1xEGB htheta2_1_1 htheta2_1_2 htheta2_1_3 Entries Mean RMS Theta angle^2 [deg^2] 3.16: Mrk421 Mrk421 θ 3.19 [32] (17) 3σ 4σ 3.6 S/B ON 1 2 AGN S/B 3% :2:1 CalOnlyx CalOnly2x CalOnly1x CalOnly2x CalOnly1x 49

52 Number of signal events [events] Class P8R1_TRANSIENT_R0 P8R1_SOURCE CalOnly_xEGB CalOnly_2xEGB CalOnly_1xEGB Combi_xEGB Combi_1xEGB log (Energy [MeV]) 3.17: Mrk421 SOURCE CalOnly1xEGB TRANSIENT CalOnlyxEGB 50

53 S/B 3 S/B ratio Class P8R1_TRANSIENT_R0 P8R1_SOURCE CalOnly_xEGB CalOnly_2xEGB CalOnly_1xEGB Combi_xEGB Combi_1xEGB log (Energy [MeV]) 3.18: Mrk421 / 51

54 Significance [σ] Class P8R1_TRANSIENT_R0 P8R1_SOURCE CalOnly_xEGB CalOnly_2xEGB CalOnly_1xEGB log (Energy [MeV]) 3.19: Mrk421 52

55 PSF OFF 3.7 LAT Mrk GeV 5- ACD Cal1MomZDir > LAT ACD ACD ACD 53

56 54

57 4 Cherenkov Telescope Array 4.1 Cherenkov Telescope Array (CTA) CTA (Large-Sized Telescope : LST) (Meddium-Sized Telescope: MST) (Small-Sized Telescope: SST) 4.1 H.E.S.S. MAGIC VERITAS 1 mcrab 14 erg s 1 cm 2 20 GeV - 0 TeV LST LST 2016 LST : CTA LST MST SST 23 m 12 m 6 m GeV - 1 TeV 0 GeV - TeV TeV - 0 TeV 55

58 but we do not expect the results described here to change in any significant way. The exact details of the sensitivity for CTA in general depend on the as of yet unknown parameters like the array layout and analysis technique of 4.1: CTA -8 s -1 ) -2 dn/de (erg cm 2 Differential Flux E Crab Nebula Synchrotron LAT - yrs (inner Galaxy) LAT - yrs (extragalactic) CTA - 00 hrs Photon Energy (MeV) 5 6 Inverse Compton H.E.S.S. - 0 hrs CTA - 0 hrs : CTA Figure 1: Differential sensitivity (integral sensitivity in small energy bins) for a minimum [31] significance of 5σ in each bin, minimum events per bin and 4 bins per decade in energy. 5σ 4 For Fermi-LAT, the curve labeled inner Galaxy corresponds to the background estimated at a position of l =,b=0, while the curve labeled extragalactic is calculated using the isotropic extragalactic diffuse emission only. For the ground-based instruments a 5% systematic error on the background estimate has been assumed. All curves have been derived using the sensitivity model described in section 2. For the Fermi-LAT, the pass6v3 instrument response function curves have been used. As comparison, the synchrotron and Inverse Compton measurements for the brightest persistent TeV source, the Crab Nebula 56 are shown as dashed grey curves.

59 4.2 LST 20 GeV LST : CTA 23 m 389 m 2 28 m f/d t /20sec (Photomultiplier Tube: PMT) GeV PMT (Q.E.) 35% 4 p.e. 0.02% F.W.H.M. 3 ns 1 p.e p.e. LST (Hamamatsu Photonics K.K.: HPK) PMT R PMT PMT - (CW) PMT PMT PMT CTA Preamp for CTA PACTA (High Gain: HG) (Low Gain: LG) 1200 Ω 80 Ω PMT 57

60 LST PMT CW circuit Focal Plane Instrument PACTA board PMT cluster PMT module 4.3: LST - (CW) R

61 4 4.4: [33] 4.3 (PMT) PMT 4.4 PMT 1. THBV3_0201JA 2. (Q.E.) (Q.E.) hν 59

62 4.5: PMT R R Q.E. 40.8% ( 4.5) nm Q.E. HPK 95% δ E δ = a E k (4.1) 60

63 a k δ α µ V µ = α δ 1 δ 2 δ n (4.2) ( ) V kn µ = α (a E k ) n = αa n (4.3) n + 1 PMT 350 V LST PMT R % 200 C ns - µs PMT 4 p.e. 0.02% R HPK % MHz PMT 3ns 3 ns 61

64 4.6: PMT R

65 R R FWHM 3ns 00V PMT 00V 850V F F noise excess factor PMT F F F F PMT F 63

66 5 CTA 5.1 CTA-LST (PMT) LST PMT PMT PMT Q.E. V op PMT PMT Q.E. PMT 5.1 PMT PMT PMT CTA Q.E. F Q.E. HPK PMT PMT 3 ns PMT R V FWHM 3 ns HPK 00 V 850 V 22 V PMT 3 ns 64

67 5.1: Q.E. PMT 0 V PMT PMT PMT 65

68 5.2: 5.3: PMT PMT ZQ3932 ZQ3672 ZQ vs. ZQ3672 ZQ V PMT V 5.5 FWHM 3.7 ns 0.3 ns 3 ns 66

69 5.4: PMT ZQ3932 ZQ3672 ZQ3672 vs. 5.5: PMT 67

70 5.6: 0% PMT 700 V PMT C 20% V 00 PMT 0 V PMT 5.7 PMT PACTA(ASIC) PACTA V PMT PACAT 160 V 06 V 5.8 FWHM 3.57 ns

71 Original Factor 3.54 (2.00) attenuation PMT 27 Ω PMT 154 Ω (76.8 Ω) PACTA 23 Ω 15 kω 69.8 Ω (0 Ω) PACTA 23 Ω 15 kω GND GND GND GND GND 5.7: PMT-PACTA 3.5(2.0) 1 5.8: 1 ns PACTA PMT R vs. ( ) V 350 b/7 Gain = a (5.1) 8 69

72 350 HPK V N PMT A V op = A 1/b (V N 350) (5.2) V op PACTA PMT R b = 4.83 PACTA HPK 5.1 PMT : PMT 00 V 3 ns V 1 V V 172 V PACTA PMT CTA-Japan LST PMT 2 (Data Aquisition: DAQ) PMT PMT DAQ 70

73 [pieces] hvoltageg4e4 Entries 2027 Mean RMS Operation voltage hvoltageg4e4mod20 Entries 741 Mean 30 RMS Attenuation factor No adjustments hvoltageg4e4mod35 Entries 1273 Mean 50 RMS hvoltageg4e4mod Entries 13 Mean 06 RMS Operation voltage [V] 5.9:

74 PMT CTA-Japan : PMT PMT PMT PMT Raspberry Pi DAC I 2 C DAQ DRS4 DAQ imac - Single p.e. - Multi p.e. - After pulse p.e. MySQL PMT 00 x 500 x 5 mm PMT 6 8 PMT PMT 5.11 PMT PMT 72

75 Filter wheels Voltage controller Measurer Controll Order System Contoroller PMT module DAQ program Data user Access Analysis program Data base Output DRS4 evaluation board 5.: 73

76 5.11: PMT 74

77 (Laser dyode: LD) NDV nm PMT LD NDV4212 LED LD [36] FWHM ps % [37] LD LD LST PMT 1 p.e p.e. 6 ND PC 5.11 DAQ DRS4 PMT FWHM 3 ns The Paul Scherrer Institute (PSI) DRS4 DRS4 (1-5GHz) 200MHz ADC DRS4 PMT DRAGON DRS GSampling/s 5 GSampling/s 1 DRS4 4 PACTA HG LG USB imac -500 mv 500 mv 75

78 図 1 電源モジュールのセットアップの様子 (2014/8/22) ADC 入力口 Raspberry Pi DAC 出力口 電源 LAN DAC ADC リファレンス OP アンプ 図 2 HV コントロール回路の部品 5.12: 2 LD 4 DRS4 TTL DRS4 DAQ C++ CERN ROOT Raspberry Pi PMT PC Raspberry Pi DAC I 2 C PMT 8 PMT PMT 0.5% 5.12 DAQ 76

79 PMT HG PACTA HG LG PMT Multi p.e. HG DRS V 1500 V LG HG 8 imac USB 4 USB PACTA LG HG : V op Multi p.e. ( + ) [p.e.] Single p.e. F 1400 V Multi p.e. Multi 2nd p.e. F F 1400, 1300, 1200, 10, 00, 950 V ( ) 6 20 V op After pulse V op Single p.e V 1400 V F 00 V 77

80 5.13: PMT 1 PMT 1 p.e. 1 1 p.e. PMT G P MT = < q 1 > eg Amp (5.3) G PMT G Amp PMT < q 1 > e < q 1 > σ 1 F ( ) 2 F 2 σ1 = 1 + (5.4) < q 1 > p.e. 0 p.e., 1 p.e., 2 p.e., 3 p.e., [34] 1 p.e LD 0 mv ns [35] 78

81 1. 1 p.e. PMT PMT σ (σ ) p.e. (5.3) (5.4) F [events] Output charge of lighted hlit_aligned Measurement Lighted Dark Entries Mean RMS Underflow 0 Overflow 0 Integral 6.084e+04 hdark_aligned Entries Mean RMS Underflow 0 Overflow 0 Integral 4e Charge [mv ns] 5.14: 1 2 p.e. 1 p.e. 1 p.e p.e. 1 p.e. 1% F (5.4) 4% DRS PMT 79

82 [events] 3 Charge of (Dark subtracted) hremn Entries Mean RMS Charge [mv ns] 5.15: Multi p.e. 1 p.e. 1 1 p.e. n p.e. p n (n 0) 2 p.e. p 2 (x) = 1 p 1 (y)p 1 (z)δ(x y z)dydz = 1 p 1 (y)p 1 (x y)dy (5.5) 2p 0 2p 0 1 p.e. 1 p.e. p 0 1 p.e p.e. 3 p.e. 4 p.e p.e. 1 p.e. 1 p.e. 2 p.e p.e. (5.3) (5.4) F 80

83 Charge of (Dark subtracted) [events] 2 Component 1 p.e. -0 (Total signal) 2 p.e p.e p.e Charge [mv ns] 5.16: [events] 2 Charge of (Dark subtracted) Component Total signal 1 p.e p.e p.e p.e Charge [mv ns] 5.17: [events] 2 Charge of (Dark subtracted) Component Total signal 1 p.e p.e p.e p.e Charge [mv ns] 5.18: 81

84 5.19: 1 p.e Multi p.e V vs. Single p.e. (V op ) F Multi p.e. p.e. Single p.e. vs. Single p.e. F Multi p.e. PACTA High Gain (HG) Low Gain (LG) HG PMT Multi p.e. vs. p.e. Single p.e V 1400 V) Multi p.e. (5.1) vs. 82

85 5.20: PMT ZQ3197 F F m σ 1 p.e. p.e. F F 2 m = (σ/p.e.)2 m/p.e. (5.6) Single p.e. 1 p.e. F FWHM ( 50% 50% ) ( % 90% ) ( 90% % ) 5.20 PMT GSampling/s PACTA HG LG LG HG DRS V 1500 V LG HG vs. HG LG 1200 V 1300 V LG HG 5.21 HG 5.22 HG Single p.e. 83

86 5.21: 1200 V 1300 V (LG) vs. (HG) 84

87 5.22: vs. HG Multi p.e V Single p.e. 85

88 5.23: vs. Multi p.e. HG 1400 V Single p.e. PACTA HG HG DRS V HG LG 5.23 HG Single p.e. 5.5 F F F PMT PMT PMT ZQ3197 ZQ

89 ZQ7532 Single p. e. F V 1400 V ZQ3197 ±6% ±3% F 2 ±4% ±1% PMT 5.4: PMT ZQ3197 ( 1500V) (±6%) 5 3 (+9%) (-21%) 5.24 ZQ3197 ( 1400V) (±2%) (+9%) (-7%) 5.25 ZQ7532 ( 1400V) (±3%) (+9%) (-4%) : PMT F F 2 ZQ3197 ( 1500V) (±4%) (+19%) (-3%) 5.27 ZQ3197 ( 1400V) (±1%) (+7%) (-3%) 5.28 ZQ7532 ( 1400V) (±1%) (+3%) (-2%) Single p.e PMT 0.2p.e. F 0.03p.e p.e. 3.4 F 2% 0.2 p.e p.e. 8 % 0.03 p.e. 87

90 [pieces] Gain of ZQ3197 at 1500V hgain Entries 140 Mean 4.667e+05 RMS 2.696e Gain 5.24: PMT ZQ3197 single p.e. 1500V Gain of ZQ3197 at 1400V [pieces] hgain Entries 91 Mean 2.911e+05 RMS Gain : PMT ZQ3197 single p.e. 1400V 88

91 Gain of ZQ7532 at 1400V [pieces] hgain Entries 91 Mean 3.065e+05 RMS Gain : PMT ZQ7532 single p.e. 1400V [pieces] 50 2 F of ZQ3197 at 1500V hf2 Entries 140 Mean RMS F 5.27: PMT ZQ3197 single p.e. F 1500V 89

92 [pieces] F of ZQ3197 at 1400V hf2 Entries 91 Mean RMS F 5.28: PMT ZQ3197 single p.e. F 1400V [pieces] 40 2 F of ZQ7532 at 1400V hf2 Entries 91 Mean RMS F 5.29: PMT ZQ7532 single p.e. F 1400V 90

93 [events] 2 Charge of (Dark subtracted) hremn Entries 0477 Mean RMS Charge [mv ns] 5.30: PMT ZQ p.e. F 2 2% ±% F ±1 2% p.e V 91

94 5.6: ZQ p.e p.e. Multi p.e. 0% ZQ3197 (1400V) F 2 Multi p.e. 0.03p.e (0.0%) 0.2p.e. 362(±8) 3 (115.3%) (98.1%) 1.24 (0.0%) 1.22 (98.4%) (291 ± 6) 3 (92.7%) 1.35±0.01 ( 9.0 %) 1.23±0.02(98.8%) 5.7: ZQ p.e p.e. Multi p.e. 0% ZQ7532 (1400V) F 2 Multi p.e. 0.03p.e (0.0%) 0.2p.e. (390 ± 12) 3 (137.8%) (97.9%) 1.24 (0.0%) 1.22 (98.4%) (307 ± 8) 3 (8.5%) 1.34 ± 0.01 (7.8%) 1.20 ± 0.01 (96.6%) 5.8: ZQ p.e. Multi p.e. 0% ZQ4504 (1400V) F 2 Multi p.e. 0.03p.e (0.0%) (98.2%) 1.24 (0.0%) 1.22 (98.4%) 0.2p.e (139.7%) (115.1%) 1.33 (7.2%) 1.23 (99.2%) 92

95 5.9: ZQ p.e. Multi p.e. 0% ZQ4568 (1400V) F 2 Multi p.e. 0.03p.e (0.0%) (98.3%) 1.23 (0.0%) 1.22 (99.2%) 0.2p.e (126.4%) (6.8%) 1.30 (5.7%) 1.21 (98.4%) 5.: ZQ p.e. Multi p.e. 0% ZQ5009 (1400V) F 2 Multi p.e. 0.03p.e (0.0%) (98.5%) 1.21 (0.0%) 1.20 (99.2%) 0.2p.e (126.5%) (5.5%) 1.30 (7.4%) 1.21 (0.0%) 5.11: ZQ p.e. Multi p.e. 0% ZQ5562 (1400V) F 2 Multi p.e. 0.03p.e (0.0%) (98.3%) 1.21 (0.0%) 1.19 (98.4%) 0.2p.e (118.6%) (98.7%) 1.28 (5.8%) 1.19 (98.4%) 93

96 Charge [mv ns] 5.31: PMT ZQ7532 Single p.e PACTA PACTA PMT PMT 500 PACTA PMT PACTA 5.1 HPK V PMT V PMT PMT 500 F vs (5.1) (5.7) Gain = a(v 350) b + c (5.7) 94

97 5.32: vs. PMT PMT generation vs. Gain curve power law index Serial No Power law index 5.33: (5.7) PMT R

98 Operation voltage (Gain=40000) Operation voltage at ICRR [V] Attenuation factor Operation voltage at HPK [V] 5.34: HPK 4 PACTA 1.0, 2.0, 3.5 PMT CTA (5.1) 5 - V (5.7) 5.33 (5.7) 5500 HPK 5.34 HPK PMT 5.35 HPK PMT (5.7) PMT vs. 96

99 Operation voltage (Gain=40000) Operation voltage at ICRR [V] PMT serial No. <= 5500 > 5500 Expected <= 5500 > Operation voltage at HPK [V] 5.35: 5500 PMT 5500 PMT HPK PMT HG LG HG 5.36 HG LG PMT HG PMT (5.7) PMT PMT HPK 1.2% F PMT Single p.e. Multi p.e. F Single p.e

100 1140 Operation voltage at ICRR [V] Operation voltage (Gain=40000) Measured with HG and LG HG only Operation voltage at HPK [V] 5.36: HG LG PMT HG PMT [pieces] Deviation of operation voltage of PMT whose factor 2.0 Measured with HG and LG HG only h1voltageg4e4dev_dummy20 h1voltageg4e4dev20_0 Entries Mean RMS h1voltageg4e4dev20_1 Entries 348 Mean 2.14 RMS (V - V exp )/V [%] op op 5.37: 2.0 PMT 3% HG LG 98

101 [pieces] Deviation of operation voltage of PMT whose factor 3.5 Measured with HG and LG HG only h1voltageg4e4dev_dummy35 h1voltageg4e4dev35_0 Entries Mean RMS h1voltageg4e4dev35_1 Entries 144 Mean RMS (V - V exp )/V [%] op op 5.38: 3.5 PMT % Multi p.e % FWHM FWHM 5.40 ( ) V 350 c f(v ) = a + b (5.8) V PACTA HG LG PACTA LG 360 MHz 480ṀHz HG FWHM PMT ns ns PMT FWHM 3 ns PMT PMT ns ns FWHM 5.5 ns 99

102 F^2 (MultiPE) hf2 Entries 917 Mean x Mean y RMS x RMS y F^2 (SinglePE) 5.39: F F 2 Single p.e. Multi p.e. 5.40: 4 PMT FWHM 0

103 [pieces] 300 FWHM time h1fwhm Entries 916 Mean RMS [ns] 5.41: PMT FWHM FWHM time vs. operation voltage FWHM time[ns] Operation voltage[v] 5.42: PMT FWHM 1

104 5.43: 4 PMT [pieces] Rise time h1rise Entries 916 Mean 2.23 RMS [ns] 5.44: PMT 2

105 5.45: 4 PMT [pieces] 140 Fall time h1fall Entries 916 Mean RMS [ns] 5.46: PMT 3

106 vs. HPK HG LG LG HG HG 2-3% 9-14% HPK DC DC PMT DC HPK DC PMT ±3% (5.1) 0.6% HPK 1% PACTA LG HG % HPK F Multi p.e. F 2 PMT Single p.e. 3% (5.6) F 2 Multi p.e. F 2 Single p.e. Multi p.e. 2.5% R F p.e. 2 F 4

107 F = PMT 5.8 LST PMT DC 5.9 PACTA PMT p.e. PMT PACTA HPK 920 V PMT 3.5 PACTA 10 V PMT 5

108 6 GeV GeV CTA Mrk % CTA DC 6

109 [1] K.A. Olive et al. (Particle Data Group), Chin. Phys. C, 38, (2014). Review of Particle Physics, 2014 [2] M. Ackermann et al., 2013, science [3] The Fermi-LAT Collabration, 2013, arxiv: [4] T. C. Weeks, Very high energy gamma-ray astronomy, 2003, Institute of Physics Publishing [5] CORSIKA Shower Images, < [6] M. Rissi, 2009, PhD thesis, ETH Zurich [7] Cherenkov Telescope Array 2014, CTA-Japan Consortium < [8] T. Takahashi et al. 2012, arxiv: [9] TeVCat S. Wakely and D. Horan [] W. B. Atwood et al. 2009, ApJ [11] The MAGIC Collaboration, 2011, 32nd International Cosmic Ray Conference [12] C. Urry, 2003, arxiv:astro-ph/ [13] M. Ackermann et al ApJ [14] A. A. Abdo et al ApJ 736 [15] A. Franceschini et al. 2008, A&A 487 [16] S. Matarrese et al. Dark Matter and Dark Energy. 2011, Springer 7