coherent WaveBurst Review Meeting 26 Mar 2009 09:00 Pacific / 12:00 Eastern
Agenda: Thursday 26 Mar 12:00 Eastern
- Burst review timeline for current papers: [minutes page]
- Reports on action items.
- Calibration tests pages:
Dial-in number: 1 800 704 9896 International dial-in: +1 404 920 6472 Conference code: 5374 2349 #
AttendancePatrick (minutes), Marco, Gabriele, Francesco, Igor, Sergey, Virginia, Irena, Stas.
- ExecComm meetings: April 16, 23
- APS: May 2-5
- LVC: June 2-5
Action item updates
- Segment lists provided for both LFS and HFS -- PJS to check.
- IY to make tables of found/missed injections and loud background for RG to follow-up, by next telecon.
- PJS to check with MZ about IY's post-proc scripts -- were they reviewed for year-1?
Calibration studyVR: In email contact with Stas over last 3 weeks discussing details of calibration tests already done (for different networks and waveforms).
- cWB tests up to 20% amplitude error and 44 deg phase error.
- Separate tables for different networks and calibration error tested. Each table report the average difference as 4.2% -- looks like a copy-and-paste error.
Action item: VR to check on the 4.2% number.
- Phase errors done by time shifting. One sample at 16384 Hz = 60microsec which is about 25 deg at 1 kHz. To get finer time shifts, up-sample injection data to 64kHz, time-shift, then down-sample and process normally.
Discussion of how to best represent effect of calibration error. FS: Mis-calibrated injections done on JW1 only, but hrss values compared to full 2cd year. PJS: Looks like differences are dominated by difference between JW1 and full year. Fluctuations in hrss for miscalibrated injections follow fluctuations in hrss for regular JW1 injections. Suggest report differences between JW1-miscalibrated and JW1-calibrated.
Action item: VR to add extra line to tables showing difference between JW1-miscalibrated and JW1-calibrated hrss.
- PJS: Are measurements limited by number of injections? FS: Smallest data set about 4 days, so 1 injection per 100 sec and ~10 waveforms gives ~300 injections per waveform and amplitude, so ~6% error. However, these small sets do not contribute much to limit. Larger sets will have much better statistics. Then Monte Carlo uncertainty negligible.
- PJS: Hughey recently circulated proposal for incorporating calibration uncertainties in ULs. Should read and comment.