Present:

     Caltech: Ben Cusack,  Ron Drever, Szabi Marka,  Peter Shawhan, Stan Whitcomb, John Zweizig
     Hobart: Steve Penn
     LHO: Dick Gustafson,  Mike Landry, Rick Savage, Robert Schofield
     Michigan: Keith Riles
     MIT: Rana Adhikari, Laura Cadonati,
     Oregon: Ray Frey, Masahiro Ito, Rauha Rahkola, Brian Stubbs
     Penn State: Mike Ashley, Patrick Sutton

S2  Status Update Reports

• Investigation teams

 
• Data quality  (K.Riles)

KR gave a brief update on the status of the DQ segment repository.  Many new studies have been completed (or are nearly completed) and will be incorporated into version 3 segments.

One point of discussion was flagging periods when calibration drift coefficients are "out of bounds". It was decided to flag periods when alpha is not between 0.5 and 2.0 (1.0 is nominal). PatrickS agreed to generate tables of bad times. He has already produced tables of times when no coefficients could be determined.

One recent piece of bad news was that a number of H1 channels had a byte-swapping bug at the hardware level during all of S2 (see note from Daniel Sigg).  Robert pointed out, however, that the PEM channels affected have not proven useful in the past in understanding H1 behavior. For historical reasons, many H1-related PEM channels are read out via H2 ADC crates.
 

• Calibration stability (M. Landry, P. Sutton)

The iteration 2 calibration is now complete! More information can be found at the calibration team home page and at Patrick's S2 SenseMonitor page. Although a careful error accounting has not been done yet, it is believed that iteration 2 is comparable in accuracy to the final S1 calibration, i.e., at the 10% level.

The final (iteration 3) calibration is expected to be ready at about the time of the LSC meeting at Hannover. The changes between iterations 2 and 3 come from incorporating measurements of electronics delays and transfer functions and are expected to be most important at frequencies above 1 kHz. The regeneration of alpha coefficients should go much faster in the next iteration because it will not be necessary to go back to the raw data. MikeL confirmed that the team believes the timing uncertainties in transfer functions that prevented coherent, multi-IFO pulsar tracking at 1.3 kHz in S1 will be much smaller after the final calibration. (Note: it is planned to use atomic clocks in the S3 run to give backup timing info / confidence.)

MikeL is now generating "stardard" S2 power spectral estimates for use in external LSC presentations. He is producing one representative curve each for H1 and H2, but expects to produce three curves for L1, since its inspiral range improved dramatically during the course of the run.
 

• Glitches (L. Cadonati, M. Ashley)

Laura provided a link to the glitches team web page and gave an overview of ongoing work. The MIT group is running glitchmon on triple-coincident H1 & H2 playground RDS data at MIT (but is having technical problems running on the LHO fortress machine to look at other data samples). The Syracuse team continues looking through L1 playground segments to understand glitch and glitchmon triggers; and is learning how to do glitchmon reruns to try varying thresholds. The Florida team continues WaveMon studies, but has also had trouble running at LHO (Note by KR: later e-mail exchanges resolved at least one of these troubles.).

MikeA reported on some calibration line tracking he has been doing,  using a straight heterodyne of ASQ. He confirms drifts in phase of the lines and wants to use those drifts to flag data quality. KR suggested comparing his magnitude/phase determinations with those recorded by Sergey's LineMon monitor (available as minute trends). Rana was happy to see a heterodyne approach being used, but warned that one should use filtering in addition.  PeterSh pointed out that MikeA was incorrectly assuming all IFO's had the same calibration line frequency. KR volunteered to send MikeA the correct values for each IFO.
 

• Violin modes (S. Klimenko, J. Castiglione)

Not present. Sergey reported by e-mail later that LineMon studies at LHO have suffered the same troubles as those for WaveMon (see above).
 
• Steady-state correlations (N. Christensen)

Not present, but Nelson sent a link to the Carleton group's glitch and correlation web page. which summarizes ongoing activity. The most recent work concerns bicoherence analysis and studies of hardware signal injections.
 
• Bilinear couplings (S. Penn)

Steve continues  work on bilinear coupling DMT code. More details on his and Vijay's recent work will be given at the Hannover meeting.
 
• Correlated inter-site environmental transients (R. Schofield)

Greg Mendell has run into a number of technical problems in creating the inter-site PEM RDS. Completion of the RDS is expected in about two weeks. The RDS will contain magnetometer, RF, HV and seismic channels mainly. No analysis results are expected before the Hannover meeting.
 
• Local environmental disturbances (R. Schofield)

Robert reported mainly on recent progress in acoustic mitigation (see June 13 minutes for background). Various decisions have been made concerning electronics racks, periscope redesign, optical table legs, etc., in preparation for S3. In parallel, Masahiro continues studies of S2 vetoes to understand PEM couplings to ASQ.
 
• Data access (P. Shawhan)

Following discussions at the last detchar, Peter put together a web page with a detailed list of the locations of all S2 data samples (full and RDS), including channel lists and applicable DMT computers with data access. KR suggested sending out an announcement of this quite useful page to the ligo-dc mailing list.

Peter also reported that the second trends now in HPSS storage at Caltech are being moved to the new SAM-QFS facility, which should allow much better access to those trends via LDAS. A number of persons have requested access to that information for the S2 run.
 

• Data set reduction (I. Leonor)

No update.
 
• Signal injections  (S. Marka, P. Shawhan)

No new feedback from the search groups on hardware injections. KR has proposed a joint detchar/asis session at the Hannover meeting concerning S2 hardware injection results (see below). Szabi offered to help organize such a session.

• Planning for the August LSC meeting at Hannover

KR presented a very preliminary agenda of the detchar working group sessions at Hannover. The agenda includes 1) summary reports to the LIGO I collaboration on Monday morning concerning S2 calibration, timing, glitches & data quality; 2) more details on these and other S2 investigations; 3) talks on new software/simulation tools; and 4) GEO / ALLEGRO presentations. In addition, there will be a joint detchar / asis session (organized by Stan Whitcomb) on how each search group would verify signal detection if nature were kind but not as kind as we would like. Another joint detchar / asis session may include talks on S2 hardware signal injections.

A.O.B.

• Rana wondered whether more DMT machines could be made available at the sites for looking at S2 data. Szabi advocated using general computing (GC) machines because they are fast, but the GC machines don't currently support DMT code. He has been working to get this available at LLO. KR urged that a temporary solution of mounting data on delaronde (in addition to decatur) be used for now. GC machines could provide a medium-term solution. The Tier 2 analysis centers (UWM, PSU, MIT) provide a long-term solution. KR mentioned the approval of a new NSF ITR proposal that provides 2.25 FTE's across the analysis centers to support DMT infrastructure.

 
• Next Detector Characterization Teleconference: Friday August 8 at 1:30 p.m. EDT