Data Analysis of TAMA

Transcription

1 Data Analysis of TAMA Nobuyuki Kanda the TAMA collaborations Miyagi University of Education, National Astronomical Observatory, University of Tokyo, University of Electro-Communications, High Energy Accelerator Research Organization, Osaka University, Kyoto University, Max-Planck-Institut fur Quantenoptik, National Research Laboratory of Metrology, Hirosaki University, Kinki University, Tokyo Denki University, Osaka City University, Tokai University, Tohoku University, Niigata University, Hiroshima University TAMA symposium 2/6/2002 Hongo, Tokyo N.Kanda Miyagi Univ. of Education

2 Data Taking History Data Taking 1 (DT1) Period actual data amount 8/6-7/1999 (one night) ~3 + ~7 hours continuous lock DT2 first Phyics run 9/17-20/ hours DT3 4/20-4/23/ hours DT4 8/21-9/3/ hours DT5 3/1-3/8/ hours Test Run 1 6/4-6/6/2001 Longest stretch of continuous lock is 24hours 50min. The interferometer can operate in daytime of weedkay. DT6 8/1-9/20/ hours with duty cycle 86% h ~ [1/ Hz] N.Kanda Miyagi Univ. of Education

3 Detector Schematics (DAQ) 5 intranet Main IF signal * 7ch L- feedback, Calibration Reference etc. IRIG-B time coded signal 20 khz trigger signal (generated from 10MHz GPS clock) VXI crate MIX BUS Interface ADC (SONY tektronix VX4244) Continuous Data Taking IF Calibration Signal Record UNIX Workstation (SUN Enterprise 450, OS Solaris 2) 2GB SCSI 2GB 2GB 2GB 2 GB * 4 Disks 1st level spool disk DLT tape system (5 tapes at once)

4 On-line analysis (1) Data acquisition system 11

5 DAQ front end Acquisition Spool & Transfer Archive signals VXI-MXI crhsun1 DAQ processes (tqstart) HardDisk HDAQ raw data /export/home*/r*f*.raw DLT ~daq/online_monitor/copyrecent NFS server /usr4/ Monitor: HdaqMon display NFS fft /import/usr4/onlmon/spool/ ~daq/hdaq_mirroring Mirror of HDAQ NFS server /data/hdaq/ MDAQ raw data /data/mdaq/ signals ADC-SCSI mdaq DAQ processes NFS NFS /import/data/hdaq/ /import/data/mdaq Monitor: ArchiveMon online analysis (IF diag., mdaq) Monitor: MdaqMon crhsun2 /import/data/hdaq/ online analysis (MF, SNR) ~daq/online_monitor/startmon EPICS print task ~daq/epics_monitor/startemon print out scp signals VME cresun1 ~daq/src/tqepics/copyedata DAQ processes (tqepics) EPICS raw data Monitor: EpicsMon

6 DAQ and pre-process DAQ system DAQ Calibration accuracy h/h ~1% Pre-Process Cut-out observation data remove unlock or tuning periods h(f) : strain equivarent spectrum calculus v(t)-> v(f) T(f) -> h(f) Analysis N.Kanda Miyagi Univ. of Education

7 Two issues of the data analysis Detector evaluation IF operation stability, gaussianity Detection Feasibility GW search Stochastic Binary Coalescence Burst Continuous N.Kanda Miyagi Univ. of Education

8 Sensitivity / Data amount, and possible issues before DT Simulation Method study DT1 / DT2 : a few night run Calibration Off-line data selection Stability, non-gaussinanity check Matched filter analysis DT3 / DT4 : more than 100 hours, operation in night Veto analysis IF diagnosis Continuous analysis DT5 : preparation of 1000 hours DT6 : 1000 hours Fast IF diagnosis Expected SNR monitor = Real time monitor, feedback to the detector operation N.Kanda Miyagi Univ. of Education

9 Calibration Drift of openloop TF at 625 Hz 1.8 Gain run12 (DAQ 2) 1.0 Sep 19 20:00 Sep 19 22:00 Sep 20 00:00 Sep 20 02:00 Time Sep 20 04:00 Sep 20 06:00 40 'measured' 'gauss fit (result: 1sigma Gain drift for night ~ 30% Calibration accuracy h/h ~ yield delta G between 16 frames/mean at time [%]

10 DAQ 2 & DAQ 3: Noise Instability 6 noise average around 1kHz run12 (DAQ 2) 9 hours run34 (DAQ 3) run35 (DAQ 3)

11 Sensitivity History of Data Taking 6 SNR 1e-18 1e-19 1e-20 1e-21 08/01 08/02 08/03 08/04 08/05 08/06 08/07 08/08 SNR SNR SNR SNR SNR SNR SNR 1e-18 1e-19 1e-20 1e-21 08/08 08/09 08/10 08/11 08/12 08/13 08/14 08/15 1e-18 1e-19 1e-20 1e-21 08/15 08/16 08/17 08/18 08/19 08/20 08/21 08/22 1e-18 1e-19 1e-20 1e-21 08/22 08/23 08/24 08/25 08/26 08/27 08/28 08/29 1e-18 1e-19 1e-20 1e-21 08/29 08/30 08/31 09/01 09/02 09/03 09/04 09/05 1e-18 1e-19 1e-20 1e-21 09/05 09/06 09/07 09/08 09/09 09/10 09/11 09/12 1e-18 1e-19 1e-20 1e-21 09/12 09/13 09/14 09/15 09/16 09/17 09/18 09/19 1e-18 tuning or unlock state 1e-19 h at 1 khz h at 100 Hz 1e-20 1e-21 09/19 09/20 09/21 09/22 09/23 09/24 09/25 09/26

12 On-line analysis (2) On-line monitor One of the monitor screens for HDAQ Obs. status, power, servo gain, averaged noise level, excess. Observation Status Dark Port Power Open Loop Gain Noise Level (935Hz) Excess (Gaussianity) Obs HDAQ Data Analysis and Monitor (ver 2.10), Run : 107, File : 1200 Start : MJD, Sep 02, 2001, Sun, 09:35:21 JST, Current : MJD, Sep 02, 2001, Sun, 11:23:29 JST N/O Time (min.) 12 Obs [V] [V rms ] [deg] 9.762e 21 [1/Hz 1/2 ]

13 Interferometer diagnosis (6) Gaussianity evaluation Excess (Gaussianity) evaluated in every 1 min. Gaussianity parameter c 2 :1 (F.A < 10ppm) 14.0%. Quiet hours Threshold F.A.: 3% 11.3%. Slightly larger Consistent distribution. Investigation with the other signals. Counts Counts c 2 =1 (F.A. <10ppm) Theoritical distribution (Simulated Gaussian noise) DT6 total Excess c 2 =0.19 (F.A. 3%) c 2 =1 (F.A. <10ppm) Theoretical distribution (Simulated Gaussian noise) Quiet period (10hours, Aug. 06 night ) DT6 total (nornalized) Excess

14 Detector evaluation: Veto Analysis HDAQ signals (DT4) N.Kanda Miyagi Univ. of Education

15 L L V(t) V(t) V(t) V(t) V(t) V(t) V(f) FFT V(t) 1 2 S1frame = Σ{V(t) } 1=3.2768s s V(t) s 2 G(f) h(f) <h(f)300hz> <h(f)700hz> <h(f)1khz> h(f) log 10Hz 300Hz 700Hz 1kHz 2 S1frame = Σ{V(t) } #! 5.4 5BH= A 5BH= A 5BH= A Hz log σ σ 5BH= A NJD 5BH= A LAJ 2=II!0 %0 0!0 %0 0 # % # & DB!0 DB%0 DB 0 DB!0 DB%0 DB 0

16 26x10-6 S 1frame [V 2 ] N : ADC 4 2 S 1frame 2 S 1frame [V 2 ] ( ) 5.6: ADC 4 2 S 1frame 2 S 1frame [V 2 ]

17 [%] AND ch2 ch3 ch4 ch run : 1kHz < h(f) > 1kHz µ h1k ±5σ h1k run ch2 ch3 ch4 ch6 AND < h(f) > 1kHz µ h1k ±5σ h1k 100[%] 5.27: veto < h(f) > 700Hz veto veto log

18 60 50 Expected SNR to binary mergers at 10kpc 2001/8/3 DT6 40 SNR /6/2 2000/9/4 DT mass of a member star [Msolar]

19 Expected SNR history for typical GW event at 10kpc away SNR /01 08/02 08/03 08/04 08/05 08/06 08/07 08/08 SNR SNR SNR SNR SNR /08 08/09 08/10 08/11 08/12 08/13 08/14 08/ /15 08/16 08/17 08/18 08/19 08/20 08/21 08/ /22 08/23 08/24 08/25 08/26 08/27 08/28 08/ /29 08/30 08/31 09/01 09/02 09/03 09/04 09/ /05 09/06 09/07 09/08 09/09 09/10 09/11 09/12 SNR SNR /12 09/13 09/14 09/15 09/16 09/17 09/18 09/ tuning or unlock state Msolar Msolar Msolar /19 09/20 09/21 09/22 09/23 09/24 09/25 09/26

20 entry by bin (calculated for 1 min.) Histogram of Expected SNR for DT6 0.5Msolar-0.5Msolar 1.4Msolar-1.4Msolar 10Msolar-10Msolar expect. SNR expect. SNR entry by bin (calculated for 1 min.) expect. SNR time expected SNR > mean - 1s Total Time for Msolar for Msolar for 10-10Msolar 80.51% 82.37% 83.32% -> Efficent operation status is more than 80%

21 GW search: Binary inspiral Matched filter One step search Hierarchical search Wavelet Resampling N.Kanda Miyagi Univ. of Education

22 (2) : ( ) : 12/8/

23 (2) ( ) : ( ) : 12/8/

24 Matched filter Detector outputs: h( t) known gravitational waveform (template) n( t) noise. Post-Newtonian Outputs of matched filter: approximation ρ ( m, m, t,... ) Sn( f ) 1 2 c 2 noise spectrum density signal to noise ratio Matched filtering is the process to find optimal parameters which realize s( t ) = Ah( t ) + n( t ) = z SNR = ρ / 2 F H ~ ~ ( ) * s f h ( f ) S ( f ) n df max ρ( m, m, t,...) m, m, t, c 1 2 c I K

25 Matched filtering analysis 52 sec Read data t overlaps FFT of data Apply transfer function Conversion to stain equivalent data ρ ( t, m, m ) c 1 2 Evaluate noise spectrum near the data Sn ( f ) F H max ρ( m, m, t,... ) m, m, t, c 1 2 Event list c I K t 25ms t t t + tc tc c 2 c c 2 c

26 DAQ2: Preliminary Result of Binary Search 10 Number of events/bin χ 2 < 2.5 χ 2 < 1.5 χ 2 < 1.0 fitting to χ 2 < 1.5 background expect N BG = 2.5 preliminary SNR 2 with SNR threshold = 7.2 (which corresponds to 6.2 kpc for Msolar event, 2.9 kpc for M solar event, optimal incident direction and polarization), 2 events survived / 2.5 expected background -> 0.59 events/hour in C.L.90%

27 Log10[Number of events] ρ / χ

28 Discussion Coalescing compact binary search of TAMA300, 1000 hours data and LISM data, are progressing. We have not observed events, which significantly exceed the threshold in both TAMA and LISM s independent analysis. Even in the case if there are no significant events, we can still obtain upper limit to the event rate in the data using e.g. Poisson statistics. With 1000 hours data, we will be able to set an upper limit events /hours c.f.: Caltech 40m : 0.17/hours (90% C.L.) TAMA DT2 : 0.59/hours TAMA DT4 : 0.020/hours

29 Wavelet 20 testing code (discrete) expected SNR evaluation (H.Yakura, N.Kanda)

30 A principle of the Resampling method for GW from binary starts To recognize GW as a sinusoidal wave, distort time axis according to the GW frequency changing. -> Re-sample ADC sampling data dt = omega(t) / omega_cutoff resampling interval : dt omega(t) : frequency prediction of GW at t (-> template of GW) omega_cutoff : cutoff frequency of resampling must be chosen less than coalescence frequency in current study, we choose cutoff as 500 Hz in Keplar Motion in this analysis ( near by 1000 Hz in GW ) constant sampling interval "resampling" points distort along the chirp freq. TAMA group / Department of Physics, Miyagi University of Education Nobuyuki Kanda

31 Check3. How many templates we need? arrival time interval S/N threshold Typical interval: t 0 ±5msec (or =itaration interval/5msec) 6x10-21 excess height in FFT x template arrival time parameter [sec] TAMA group / Department of Physics, Miyagi University of Education Nobuyuki Kanda

32 18 Continuous GW from 1987A remnant (K. Soida)

33 19 (K. Soida)

34 New issue : Coincidence Coincidence between two or more detectors! Event candidates list comparison coherence of ρ(t) = (signal GWtemplates) constraint of arrival time, mass assumption of waveform correlation of full time series data h(t) N.Kanda Miyagi Univ. of Education

35 TAMA-LISM coincidence 21 Coincidence between LISM (20mIF in Kamioka mine) and TAMA300 Distance: 220 km (maximum delay for arrival time : 0.73msec) LISM: h ~ [1/Hz] 8/1-8/23/ /3-9/17/2001 total 777 hours

36 TAMA-LISM Analysis Algorithm TAMA LISM data reading data reading Matched filter max ( ρ( m1, m2, t m m t 1, 2, c tc tc c 2 c c 2 c )) t t t + t 3.27 c s Matched filter max ( ρ( m1, m2, t m m t 1, 2, c tc tc c 2 c c 2 c )) t t t + t 3.27 c s t TAMA event list LISM event list 2 t M, η, ρ χ, M, η, ρ, χ ctama tama tama tama tama keep the events in the common lock parts 2 clism lism lism lism, lism TAMA event list LISM event list for common lock parts for common lock parts Data length analyzed coincident event search ~ 121 hours

37

38 Results of coincident event search Results of onestep search for common lock parts TAMA LISM events events t c After -veto 31 events tc, M, η After -veto 3 events tc, M, η, ρ After -veto 0 event

39 LIGO-TAMA International cooperation How about combined performance? stability fake rate estimation Where is promising search? mass region kind of sources N.Kanda Miyagi Univ. of Education

40 Summary We prepared, and developed IFGW detector analysis issues: IF evaluation from the view of event detector, Event search (observational limit, experience of real data) We proceed to realistic event detection: Coincidence N.Kanda Miyagi Univ. of Education