Brief summary of decisions during & issues needing to be addressed
after October 24 E2 teleconference.


Present:

Carleton: Christensen
CIT:      Anderson, Drever, Marka, Matone, Mours, Yamamoto, Zweizig
Florida:  Klimenko, Whiting
Loyola:   McHugh
LHO:      Butler, Landry, McCarthy, Raab, Savage, Sigg, Whitcomb
LLO:      Daw
LSU:      Johnson
Michigan: Gustafson, Riles
MIT:      Adhikari, Fritschel, Shoemaker, Sylvestre, Zucker
Oregon:   Rakhola, Schofield, Strom

Reduced data set:

  * There were strong objections to any decimation of channels
    in a reduced data set, both because of skepticism that
    decimation could be implemented reliably, especially in
    the remaining time, and because some investigators believe
    they will need the full bandwidth (if not more) of certain
    channels. 

  * We will have approximately 0.5 TB of disk space on fortress
    for reduced E2 data after the run. This is less than what
    is needed to store one week of data with all of the channels
    requested so far. There is also a tape robot, and it was
    suggested we should get accustomed to analysis from tape
    on fortress, since that will be necessary in the future.

  * Team leaders are asked to send specific data channel requests
    to David Strom (strom@bovine.uoregon.edu) by the end of this
    week (with a cc to me). The agenda page of the meeting:

  http://www-mhp.physics.lsa.umich.edu/~keithr/lscdc/agenda2_e2.html

    contains links to a proposed baseline set of reduced data 
    channels and to the current channels in the Hanford data stream.

    Teams should specify by name the channels to be added and 
    any dependence of that channel selection on IFO configuration
    or on time. For example, we may decide to generate a generic
    reduced data set for every group for the entire uptime of 
    the interferometer, regardless of configuration (e.g., 70% of 
    a week) and an "expanded reduced" data set (including some 
    test points not available in raw data frames) for one or two 
    shifts when the IFO is locked stably in a 1-arm or 1-arm+PRM
    configuration. 

    Some teams interested in only a handful of channels or needing
    only a very short period of data may wish to request a dedicated
    reduced data set for their post-run analysis.

  * Once the channel requests are in and the data volumes tabulated,
    the results will be posted and a proposal made for action,
    which may involve defining two (or more) reduced data sets
    for general use.

Specific IFO configuration requests or invasive measurements:

  Tidal group:

    Full day of stable 1-arm running with alternation between
    X and Y arms (2 hours each suggested, but more thought will
    be given to this)

  Correlations group:

    Hours of stable 1-arm running

  Lock-loss group:

    Period during one shift to try modulation of laser intensity

  Seismic group:

    1-arm locks most useful

  Stationarity group:

    No special configuration

  Calibration stability group:

    Would like time with both arms locked to attempt swept sines.
    Ideally would like 1 swept sine per day and a dedicated 
    half-shift early on to work out the algorithm (level settings,
    repetitions, resolution, etc.)

    Would also like to inject 20 calibration lines continuously
    during the week on the input and end test masses at frequencies
    above and below unity gain frequency.

  Gaussianity group:

    No special configuration

  Angular fluctuations group:

    Stable 1-arm running with wave front sensing

  Impulses/transients group:
  
    Can do some useful work with only laser and mode cleaner locked.
    Would like stable 1 arm running to look for suspension wire
    releases.

  Bursts/chirps group:

    No special configuration

  End-to-end timing group:

    No special configuration

  Data integrity group:

    No special configuration

  Line noise group:

    No special configuration

  Data compression group:

    No special configuration

  Frequency noise group:

    Both 1-arm and full-IFO running. Would like to inject noise
    at test points. Request 2 dedicated shifts, 1 for each
    configuration.