LIGO Engineering Run Periodic Source Upper Limit Task
Draft Proposal Outline

Mike Zucker and Stuart Anderson

12 February, 2001

A. Mission priorities (we publicly assert we will accomplish in this run):

1.      Search for GR from radio pulsars with known EM pulse phase.
        Boxcar periodograms are formed for a small number (order 10?)
        of nearby pulsars using their known ephemerides.  Depending on
        how close we are to design strain sensitivity, a limit on the
        Crab pulsar may be astrophysically significant.  An additional
        component of this search is included under Item 2.b. below.

        - Doppler/AM modulation functions for target list.
        - Efficiency analysis based on added fake signals
        - False alarm analysis with respect to periodic
        detector backgrounds (see below).

2.      Search for unknown "strong" sources (emitting at fixed frequencies
        in their own rest frames) and characterization of
        instrumental artifacts. This will involve Fourier analysis of
        detector strain and selected auxiliary channels (e.g., laser,
        suspension, acoustic, magnetic), at successive levels:

        a. Short-duration transforms with no need for barycentric
        Doppler/AM correction (simple CW line catalog in lab
        frame(s)) to expose gross narrowband environmental interactions,
        inter-site correlations and "deadbands" where subsequent
        analyses are not promising because of interference.

        - Lab-frame line database for both detectors at ~mHz
        bandwidth, incl. calibrated local smoothed background levels.
        - List of inter- and intra-site correlated lines.
        - Transfer function estimates for auxiliary channel/strain channel
        correlated lines.
        - Identification of harmonic series and sideband structures with
        estimation of nonlinear coupling parameters.
        - Upper limit for all-direction GR flux.

        b. Fixed-direction, "medium" duration (within contiguous locked
        stretches) Fourier search. The duration will be chosen by the
        more restrictive of processing, sky patch size, and continuous
        locked duration limits. Time-domain
        resampling will be done for specific directions including
        the GC, Crab and other pulsars in the list from Task 1., or
        other promising directions (up to some reasonable number TBD).

        - Doppler/AM modulation functions for target directions.
        - Efficiency analysis with fake signals added.
        - Veto/accidentals analysis demonstrating rejection of
        spurious lines by successive cuts (Rx Doppler/AM, local
        channel anticoincidence, inter-site coincidence).
        - Upper limits for CW flux from selected
        direction catalog.

        c. Stack-slide and/or Hough transform  analysis on the
        products of 2.b. to improve rejection of local artifacts,
        broaden sky coverage, and increase search depth.

        - Doppler/AM search parameter metric, search grid(s), and
        - Efficiency analysis using added fake test signals.
        - Further statistics on accidental rates vs. additional physics cuts.
        - Upper limits for source flux from (broadened) selected
        direction catalog.

3. Develop efficient means for fabricating and injecting fake periodic
signals directly into the detector calibration system, preferably in a
"blind" fashion.  These signals will have to simulate astrophysical
signals (phase and amplitude evolution) but with "random" parameter
selections.  While this capability is implicit in the efficiency
assessments listed above, we felt it best to identify it as a distinct
task to emphasize the challenge aspect.

B. Possible additional analyses we could consider (would be great if...):

4. Additional search for "simple" periodic source motions with
sideband matched filtering on transform products from task 2.

5. Limited search through spindown parameter space on transform
products from task 2.

6. Search for LMXB sources with known directions and restricted frequency
 bands. Methods TBD.

7.  R-mode emissions. Methods TBD.

8.  All-sky search.  Methods TBD.