Status of GEO burst analysis efforts Ik Siong Heng for the GEO burst group
Outline • Introduction - who we are - GEO’s S3 run • S3 analysis - goal: to have list of GEO S3 burst triggers available ASAP - have GEO++ implementation of TFClusters (not LAL compliant) - preparing to use TFClusters and Excess Power pipelines
Introduction • There are about 5 core members of the GEO burst group • Current task is to analyse data acquired by GEO600 during S3 - GEO600 was running from 5th to 12th of November - duty cycle was 95.2% • Chosen to use TFClusters (LDAS) and Excess Power (lalapps) pipeline
Typical sensitivity for GEO S3 first week run Sensitivity has improved since!!
Issues for S3 • using Burst DSO to run TFClusters - get data to CIT and visible for LDAS done for GEO S1 data - perform injections, hardware and software • Also going to run Excess Power - installed on a machine in Cardiff • Already have HACR pipeline but not LAL compliant - HACR is GEO++ implementation of TFClusters (with some minor differences) - produces GEO triggers for S1
Data issues • GEO S3 data will be transferred to CIT • Steffen Grunewald (AEI) and David Churches (Cardiff) are contact poits for GEO data transfer - GEO S1 data has already been transfer to Caltech - might use LDR if stable, otherwise, rsync - waiting for h(t) to be generated • calibrated h(t) is in double precision (LAL: single precision)
TFClusters pipeline • I.S. Heng visited Caltech to work with Julien Sylvestre on running TFClusters on GEO data • Some success: - have run TFClusters on GEO S1 data - cast S1 data into single precision - performed injections and determined efficiency - GEO S3 data will be in the same format • ready to start processing GEO S3 data
TFClusters pipeline test inject sine gaussians into S1 data, f 0 from 0.5 to 3 kHz, from h peak 10 -15 to 10 -18 , 10 injections at amplitude per 300 second segment 10 -17 50% efficiency (threshold level) h/ √ Hz GEO S1 noise curve 10 -20 10 3
Excess Power pipeline • lalapps excess power code built on a machine in Cardiff (explorer) • Carlo Nicola Colacino and I.S. Heng visited UWM to work with Patrick Brady and Saikat Ray Majumder • successfully cast GEO h(t) data (double precision) into a single precision number - high pass filter with LAL functions to reduce dynamic range of data because casting into single precision • ready to start processing GEO S3 data
Excess Power pipeline without high pass filtering (double precision) 10 -14 h/ √ Hz with LAL high 10 -16 pass filtering (single precision) 10 -18 10 -20 10 -22 10 3
Hardware injections • Martin Hewitson and Harry Ward have been upgrading hardware • I.S. Heng has changed Inspiral injection software to inject Burst waveforms (sine gaussians) • We are aiming to co-ordinate at least 1 set of injections with LIGO detectors • Ken Strain is the contact point on this co-ordination
Vetoes • R. Balasubramanian is in charge of running GEO++ monitors to for veto investigations - GlitchMon - PowerTracker - LineTrackerMon - HACR • Started looking at GlitchMon and PowerTracker - gives an impression of glitch statistics of data
Multi-detector analysis • GEO600 has different sensitivity band (sweet spot) and orientation from LIGO detectors • Orientation issues - coincidence windows - amplitude cuts • Sensitivity band issues - frequency cuts • Perform Monte Carlo simulations to get some answers
Summary and discussion • GEO burst group is focused towards analysing S3 data • Two LAL compliant pipelines, TFClusters and Excess Power, are ready to process GEO S3 data • Many interesting and critical issues for multi- detector analysis need to be addressed
Recommend
More recommend
