E11 Investigations:

• Calibrations (Mike Landry, Gaby Gonzalez, Brian O'Reilly)
  Final E11 calibration point functions are available, but the SenseMonitor alpha/beta trends need to be regenerated for better precision. The H2 calibration was redone at the end of the run, following a fix of photodiode saturation.
  
  
• Glitches (Katherine Rawlins, Laura Cadonati, Erik Katsavounidis)
  Katherine summarized glitch group work on E11 triggers. She has tried looking at loud triggers reported by the online excess power ETG. Many can be explained by lock losses or the rogue dust monitor (early part of the run). Two triggers coincide with airplanes, but there may be no causal relationship. She also looked at CorrPower triggers and found one extreme outlier which too coincided with an airplane. Again, causality is not obvious. Much discussion ensued on indirect ways the plane could have caused a series of glitches and other artifacts seen in auxiliary channels during the interval containing the CorrPower trigger. Lindy Blackburn has looked at the significance distribution of KleineWelle triggers for each day of the run and finds the expected improvement in H1 behavior after day 2. The distributions for H2 are less clear cut, with some apparent degradation late in the run. For both IFO's, the distributions look much better than for the S3 run. Natalia Zotov has been studying correlations of glitches seen by PTMon on AS_Q and auxiliary channels.
  
  
• Environmental Disturbances (Robert Schofield)
  Robert first described a couple of problems found by David Strom while on scimon shift. 1) the infamous dust monitor left on during the first two days near the H2 dark port table and 2) a 260 Hz peak in H2 correlated with the radio PEM channel. The source of the latter was later shown to be a mistuned modulation frequency that led to 260 Hz beats.
  Since E11, Robert has been trying to understand large variations in apparent acoustic coupling strength to the IFO AS_Q channels. For example, why do relatively smooth airplane noise levels lead to glitchiness? He has found that coupling at the REFL port varies by a factor of five, with variations correlated with drops in arm power as small as 2%. Robert is also studying magnetic & acoustic couplings at the AS_Q photodetector demod boards.
  
  
• Other burst group studies - Trigger distributions (Alessandra Di Credico)
  Alessandra has tried looking at various distributions associated with the online excess-power triggers. She finds that most triggers have unphysically long durations, although their peak times often do coincide with an identifiable glitch in AS_Q. The frequency distribution peaks near the upper limit of the search band, suggesting a filtering problem. Some of confidence levels / significances of reported triggers are also unphysical. Kipp Cannon and Saikat Majumder will be rerunning over the E11 data after tracking down the current problems.
  
  
• Pulsar group studies - Quarter-Hz lines? (Vladimir Dergachev)
  Vladimir tried looking in the E11 H1 data for the same 0.25 Hz comb seen in the S2 and S3 runs. Because the E11 run is much shorter than S2 or S3, he didn't expect to see a strong effect and doesn't. But there is some evidence from folding the spectrum on itself in 0.25 Hz bands over the run that there is an excess of power at 0.25 Hz harmonics. In addition, there is also evidence of structure with 1/16 Hz and 1/8 Hz folding, as well as for 0.53 Hz folding (suggested earlier by Robert).
  
  
• Stochastic group studies - H1-H2 coherence (Peter Fritschel)
  Peter reported that although the Stochastic Group's automated machinery is not quite ready for looking at E11 data, he did try running the DTT on the longest double-coincidence lock stretch from the run, computing a 10,000-average coherence estimate (0.25 Hz bins). The expected level of coherence from uncorrelated noise is about 1.E-4 for this number of averages. Most of the 40-500 Hz band looks consistent with no correlation, but there are anomalies near 55 Hz (Robert suggested compressors), 59 Hz (Robert suggested chiller compressors), just below 180 Hz (Robert thinks not a 3rd harmonic of 59 Hz), 221 Hz (PSL periscope likely), and near 267 Hz and 280 Hz (not understood). The sidebands on 120 Hz and 180 Hz due to stack mode resonances look much better than in S3.
  
  
• Data Quality (John Zweizig, Peter Shawhan, Keith Riles)
  John has started an E11 investigations web page and started looking at DMT trends from the run. Peter will be compiling all lock losses and 10-second intervals preceding lock loss to make DQ flags. Keith has begun compiling other DQ info and hopes to get a version 1 DQ segments release out next week.
  

DMT Software Status:

• John Zweizig: Infrastructure status
  John has fixed a problem that caused spurious linefeeds to fill up the common log file used by online DMT monitors.
  
  
• Patrick Sutton: Calibration infrastructure
  John reported on Patrick's behalf that the EasyCalibrate E11 calibration files were regenerated and that SenseMonitor has been fixed to once again return calibration line amplitude seen in AS_Q.
  
  
• John Whelan / Mark Cenac / Brian O'Reilly: StochMon
  John reported that they are trying now to implement EasyCalibrate, but have had trouble getting DMT version 2.10 up and running under linux.
  
  
• Ken Ganezer: seis_blrms
  The new version of seis_blrms was tested during E11, and data dropouts seemed to be correctly flagged. Working now on flagging the filter ringdown period following a data dropout. Will work next on a circular buffer to avoid losing strip chart history in DMT viewer when monitor is restarted.
  
  
• Sergey Klimenko: BurstMon
  Looking at detection sensitivity estimates. For ideal Gaussian noise, expect ratio of hrss_50% to spectral strain density to be about 5.5 and observe 7-8 for E11 data which is not too different from S2. Don't yet understand floating point underflows that seem to bog the program down. (JohnZ is helping to track it down.)
  
  
• Giovanni Santostasi: PulsarMon
  Giovanni reported by e-mail that he is trying to understand why DMT viewer objects get garbled in the online but not offline version of his program.
  
  
• Rauha Rahkola: SuspensionMon
  No report.
  
  
• Duncan Brown: InspiralMon ( note from Duncan , sample FOM files: plot1   plot2   xml1   xml2 )
  Duncan is working on a server system to provide xml files from online search code to DMT monitors that can display results to the DMT viewer. He showed sample plots from E11 based on the inspiral monitor, where the DMT code displaying the results is known as InspiralMon. At the moment, the latency can be as large as 2048 seconds, but Duncan is working to shorten it. KR suggested providing frequency information about inspiral triggers to give people in the control room guidance on artifacts that cause high inspiral rates. Duncan will come up with a scheme to mark clearly on displays which time intervals have not yet been analyzed, to avoid confusion.