Note from Stan Whitcomb to Astrophysical Search Groups - August 11, 2003 Last week I sent out an email inviting each search group to participate in a joint DetChar/ASIS session devoted to exploring how we might deal with a possible detection of GWs. I have put together a rough scenario of a possible outcome of from the S2/S3 runs for each of the groups. My goal was to give each group a tempting detection, but one with some sort of "problem", something that doesn't quite match expectations. What I would like to ask each of the groups to do is to think about the scenario and prepare a brief outline of what you would do to present at the session, then be ready for discussion. Are there additional analyses that you would do to confirm or disprove the possible detection? Is my scenario too iffy to contemplate claiming a detection no matter what additional tests you might do? Would it make any difference if one of the LIGO sites had suffered a natural disaster (earthquake, tornado?) and the availability of funds for repairing it was uncertain? Bruce and Keith have given us a total of 90 minutes for this exercise. I propose that each group's response to the scenario aim at 10 minutes, and that we assume that setting up the scenarios, questions and discussion will fill the remaining time. Please look at the scenarios. If there is something that is unclear about what I have outlined for your group, please let me know and we can try to fix it together. Thanks for your cooperation. stan *************************************************************************** Pulsar group: In an unbiased search (i.e., searching over a range of frequencies and f-dots) aimed at the galactic center, the S2 data show a possible signal at an h of 3e-24 at a frequency of 380 Hz in the L1 detector, the level of significance is not overwhelming, about 80% confidence level. H1 and H2 have poorer sensitivity, are consistent with the L1 signal, but don't add significantly to the confidence. If the best estimate of f-dot is attributed entirely due to slowing due to gravitational waves, the distance to the (presumed) neutron star source is 300 pc. S3: Use of a template with the observed f and f-dot in the s3 data shows no evidence of a signal. However, if the same f-dot and an f which is increased by 0.7 mHz are used. L1 sees a signal of 2e-24 at a confidence level of 90%. H1 sees a signal of 1.5e-24 at a confidence level of 85%. H2 is consistent, but with a higher noise level that does not add to the confidence in a significant way. *************************************************************************** Stochastic group: Analysis of the S2 and S3 data gives the following result: S2: L1-H1 Omega = 0.3 +/- 0.2 (error bars are 90% confidence) L1-H2 Omega = 0.2 +/- 0.3 S3: L1-H1 Omega = 0.2 +/- 0.15 L1-H2 Omega = 0.0 +/- 0.2 In both runs H1-H2 appears to be dominated by instrumental correlations. One deviation from the "expected" result is that the spectrum of Omega(f) is not consistent with flat, but could fit a power law between f^{0.7} and f^{1.4}. *************************************************************************** Inspiral group: Nothing in the S2 data, but S3 has the following "event". The strongest trigger seen in the L1 search is a rho of 18 event. It is seen with the highest rho in a BH template with masses of 6.6 and 4.8 solar masses. The distance estimate is 8 Mpc. It passes the chi^2 test with flying colors. The next highest rho during the two runs is 13. At the same time H1 and GEO were down. H2 sees the event but it is much less gold-plated than L1. It shows a rho of 7 in the same template as the L1 event, but the maximum rho comes in a different template, one with BH masses of 7.8 and 6.2 solar masses. Both just barely pass the chi^2 test. The time delay between the H2 and L1 events is between +130 ms and -60 ms depending on which template in H2 you use for the coincidence. The amplitude as measured by H2 is 1.25 times that measured by L1. *************************************************************************** Burst group: There are two interesting events. The first is in S2. It is observed in H1, H2, and TAMA; L1 was down. H1 and H2 have clear correlation, with a duration of ~15 ms and a central frequency of 750 Hz. The amplitudes are consistent between H1 and H2. TAMA sees the event, but at only 60% of the LHO amplitude. Best fit time delay between H1 and H2 is 0.3 ms, but the data would also be consistent with a -1.6 ms (~ one period) time difference. The time delay to TAMA is approximately 12 ms. The circle on the sky defined by the LHO-TAMA time difference passes through the galactic plane at 20 degrees and 35 degrees for the galactic center. This is the only coincidence between TAMA and 2 LIGO detectors, and time shift analyses of H1-H2-TAMA data (100 time shifts in steps of 10 seconds) find only two accidental coincidences. The second event is in S3. This event is observed in H1, H2 and L1; TAMA and GEO were off. This event lasted ~20 ms with a central frequency of 550 Hz. The two Hanford detectors show a clear crosscorrelation with -1.8 ms best fit for the time delay, but consistent with +0.3 ms. The ratio of h determined by H1 vs H2 is 0.8 +/- 0.15 (1 sigma). H1 and L1 show a good cross correlation, with a time delay of 8 ms. The amplitude ratio H1/L1 is 1.1 +/-.15. The circle defined by the LHO-LLO time delay does not intersect the galactic plane. Time delay analysis (LHO vs LLO, 100 time shifts in steps of 10 sec), find 13 accidental events.