Status of coalescing binaries search Status of coalescing binaries search activities in Virgoactivities in Virgo

GWDAW 11GWDAW 1118-21 Dec 2006

Leone B. Bosi INFN and University of Perugia (Italy)

On behalf of the Virgo Collaboration

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 2

Introduction on Coalescing Binaries Introduction on Coalescing Binaries activities:activities:

The Virgo CB group tasks:• Binary Neutron Stars search

– Two pipelines up and running (online and offline)– Detector non-stationarity still significant– Vetoes

• We are working to extend to:– non spinning Black Hole– Spinning binaries– High mass ratio binary system

• Network Analysis– Small scale real data exchange between LIGO and Virgo– Coherent analysis– Timing accuracy

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 3

Weekly Science Runs (WSR)Weekly Science Runs (WSR)

• WSR stands for Periodic Weekend Data Taking• Collect “Science mode” data without any experiment performed on the

detector• The interferometer run in recycled "science mode“• Take data in controlled conditions to exercise data analysis procedures• Give some feedback to commissioning• Only "calibration" and "hardware injection“ are allowed (CB and Burst

signals).• On-line data analysis is performed running the Multi-Band and Merlino

CB Analysis processes.• ITF status and behaviour monitoring• Off-line analysis

Sep.Sep. Oct.Oct. Nov.Nov. Dec.Dec. 20072007 Jan…. Jan….

11 22 33 44 55 66 77

WSRsWSRs

@2006 @2006

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 4

Data Quality: WSRs sensitivityData Quality: WSRs sensitivity

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 5

horizon distance and sensitivityhorizon distance and sensitivityA

vera

ged

Ho

rizo

n f

or

NS

NS

[M

pc]

WSR1: Duty cycle 87.7% - Longest lock 15 h

WSR5: Duty cycle 64.2% - Longest lock 10 h

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 6

Horizon distance and environmental Horizon distance and environmental conditionsconditions

• Sea and Wind activities affect horizon

• These are micro-seismic conditions

Ho

rizo

n N

S-N

S o

pti

mal

(S

NR

=8)

[M

pc]

Sei

smic

act

ivit

y [a

.u.]

Sei

smic

act

ivit

y [a

.u.] Sea: RMS 220 mHz–1 Hz

Wind: RMS 30 mHz-100 mHz

- Horizon 1.4 1.4 NSNS- Horizon 1.4 1.4 NSNS

- seismic activity- seismic activity

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 7

Triggers amplitude and detector Triggers amplitude and detector behaviorbehavior

• CB trigger rate correlated to amplitude of laser frequency line injected at 1111 Hz, measuring the coupling of frequency noise to the dark fringe

WSR2 events SNRvsTimeWSR2 events SNRvsTime(Mbta/B2 veto)(Mbta/B2 veto)

Pr_B1_ACp_1111Hz Pr_B1_ACp_1111Hz Dark FringeDark Fringe

sn

rs

nr

cou

nt

cou

nt

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 8

Trigger rate and detector behaviorTrigger rate and detector behavior

• Correlation between CB trigger rate and micro seismic peak

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 9

WSR: CB hardware injectionsWSR: CB hardware injections

• Hardware Injections– This activity is performed in the first and last run night– CB signals and Burst signals are injected– A single Taylor CB signal with [1.39;1.47]Ms and fl=50

• Analysis parameters– The template bank covers the [0.9-3] Ms space– 98% minimal match– Threshold SNR=6 and Chi2 veto– Data clustered with 10ms time

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 10

WSR Hardware Injection: WSR Hardware Injection: parameters reconstruction WSR 5 parameters reconstruction WSR 5

(e.g. Merlino)

Det.Distance Accuracy: Mean=0.87 - Stdev=0.06

Chirp Mass accuracy:

Err < 0.02 %

Chirp Mass Det. Error:

< 0.02 %

• But the expected distance was 1.08 Mpc

• Mbta and Merlino provides the same results

• Calibration error due to failure in actuation electronics, fixed later on

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 11

WSR Hardware Injection: WSR Hardware Injection: parameters reconstruction WSR6 parameters reconstruction WSR6

(e.g.Mbta)(e.g.Mbta)

Mean: 19.9Mean: 19.9

RMS: 1.8RMS: 1.8Mean: 0.99Mean: 0.99

RMS: 0.07RMS: 0.07Mean: 1.242Mean: 1.242

RMS: 0.004RMS: 0.004

Mean x: 14.1Mean x: 14.1

Mean y: 14.2Mean y: 14.2

RMS x: 1.3RMS x: 1.3

RMS y: 2.14RMS y: 2.14

Mean x: 19.9Mean x: 19.9

Mean y: 8.08Mean y: 8.08

RMS x: 1.88RMS x: 1.88

RMS y: 7.97RMS y: 7.97

Mean: -0.066Mean: -0.066

RMS: 0.31RMS: 0.31

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 12

WSR Hardware Injection: WSR Hardware Injection: Detection and Chi2 vetoDetection and Chi2 veto

WSR6 WSR6 (Mbta)

rr22

SNRSNR SNRSNR

cou

nt

cou

nt

WSR5 WSR5 (Merlino)

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 13

VetoVeto

A posteriori:• Remove high SNR triggers

– Power variation in cavities: e.g. B2• We follow the mean and RMS behavior

of Pint_B2 • Comparing RMS with a reference value, we flag a period as: beginning of noisy period

Veto a priori data quality cuts:

• Monitor for saturation in coil current in NE and WE towers

• Monitor for picomotor• Monitor SSFS saturation• …

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 14

WSR1 and B2 Veto efficiencyWSR1 and B2 Veto efficiency

• The periods tagged as noisy by "B2 veto" are well correlated to high SNR triggers

• As when applying an offset to these periods the rejection efficiency decreases

we are not randomly we are not randomly

excluding eventsexcluding events..

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 15

B2 Veto applied on WSR1 dataB2 Veto applied on WSR1 dataThe original data setThe original data set

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 16

B2 Veto applied on WSR1 dataB2 Veto applied on WSR1 dataThe data set after the B2 vetoThe data set after the B2 veto

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 17

Network Analysis: Timing AccuracyNetwork Analysis: Timing Accuracy(poster)(poster)

Study how timing accuracy can be improved considering a reference time at some intermediate frequency.

• Source direction reconstruction by a network of detectors depends on signal timing accuracy at each site

• In a template bank analysis, timing uncertainties are dominated by errors due to signal/template parameter mismatch

• This error accumulates across the frequency band as signal timing is usually taken at the end (max frequency)

"timing accuracy variation between H1 and Virgo

as a function of the reference frequency fs"

Poster Session D.GrosjeanD.Grosjean

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 18

Network Analysis: Coherent AnalysisNetwork Analysis: Coherent Analysis(poster)(poster)

• In order to start a gravitational wave sky-map, it is important the accuracy in the source parameters estimation

• There are two different methods to analyze network data and extracting stellar parameters:

2) “construct” an ideal detector equivalent to the network, to which each real detector coherently contributesCoincidence Analysis

1) Compare event list, searching from a compatible ones on each detector

Coherent Analysis

How can a coherentcoherent follow-up improve

source direction reconstruction?

Poster Session S. BirindelliS. Birindelli

Leone B. Bosi – INFN Perugia - On behalf of the Virgo Collaboration 19

ConclusionsConclusions

• Binary neutron stars searchBinary neutron stars search– Mature pipelines – Use WSR data to test and refine the analysis

procedure– Understand how detector non stationarities affect

the analysis– Work on data quality and vetoes– Pipelines now routinely run online, providing useful

information• Now moving to extend the searchNow moving to extend the search

– Wider parameter range– Share the load between the two pipelines

• Prepare network data analysisPrepare network data analysis