Review Committee Meeting Monday 10 July 2006 09:00 PST / 12:00 EST
Minutes: Monday 10 July 2006 09:00 PST / 12:00 EST
Agenda and Contact Info
- Continue discussion of comments on the draft paper for
the S2-S3-S4 GRB-GWB search
[ PDF ]
- Continue discussion of SGR1806-20 search
[ HTML ]; discuss sensitivity of the search to sine-Gaussian
[ HTML ]
AccuConference teleconferencing service:
Phone: 1-800-704-9896, participant code: 038621#
International callers ++1-404-920-6472 with same code
- GRB Search
- More results from new simulations. Ratios of linear UL : circular UL
show a factor of 5 for H1 : H2; a different ratio for H1 : L1.
Not clear if the factor of 5 is expected for any sky position or if this
is a result of averaging over GRBs (coming from all over the sky).
Would be nice to understand.
- S4 ULs have been recomputed. Distributions look about the same
(though ULs are different).
- Isabel has made progress on simulations for statistical tests.
Should have something to report next week.
- Jolien has checked delta-T and Fave for the GRBs in the technical
document. Hopes to check the computation of the cc value.
- SGR Search
- So far have been injecting monochromatic CW waveforms ... now inject
sine Gaussians. Concerned about power leakage.
- First plot: amt. of data rejected vs. veto threshold and time.
This is for no injections. Always less than 10%.
- Veto tuning in second plot: sensitivity vs. veto threshold for
monochromatic waveforms. Sensitivity increases as threshold is tightened.
- Last plot: sine-Gaussian injections. Q=10^6 is approximately
monochromatic ... reproduces earlier results. Note that hrms is a
measure of amplitude. Would be better to plot hrss (a measure of
power) to see how much power is lost to/vetoed by neighbouring bands.
Replace hrms with hrss in plot
- Note: this plot has injections after veto. Would be
interesting to see it before veto. Would expect veto only
to adversely affect very loud (and easily detectable) signals unless
they have small Q. Show this.
- Possibly generalize injections to have both AM and FM.
- MatApps review: Should there be some requirement for an ETG before
time is spent on reviewing it? Some demonstration of orthogonality or
- Peter: Only one UL will be presented; other searches only detection
searches. Don't want to put up too many hurdles to developing new
- Problem: there are 7 internal reviews in progress or requested.
This is more that we can handle. Need to balance need balance the
desire not to stifle innovation vs. not repeating nearly-the-same
analysis. This is a issue for the Burst Group.
- Immediate solution: external reviewers will focus code review
efforts on only the most vital searches (e.g., Brian will be part of
the WaveBurst reviewers; may have to stop participating in the BlockNormal
code review unless it is simply updating). Not enough external reviewers
to be part of every internal code review group.
$Id: minutes-2006-07-10.html,v 1.3 2006/07/11 02:01:03 jolien Exp $