LSC-Virgo Burst Analysis Working Group


Burst Group Home
ExtTrig Home
CBC Group Home
LSC, LIGO, Virgo


'How-to' docs
[2008, 2007, earlier]
Talks/Posters [pre-wiki]
Paper plans
White papers


Analysis projects
Old notebook [General, S2, S3, S4, S5]
Virgo workarea
External collabs


Main review page

Review Committee Meeting Monday 11 June 2007 08:00 Pacific / 11:00 Eastern

Minutes: Monday 11 June 2007 08:00 Pacific / 11:00 Eastern

Agenda and Contact Info


  1. Continued review of GRB070201 search
    • Summary [ HTML ]
    • Latest Version of Paper [ PDF ]
    • Isabel's Summary Page [ HTML ]
    • Code to check cross-corrrelation statistic [ HTML ].
      Documentation [ PDF ].
      Upper limit confidence belt [ PDF ].
      Cross-correlation distribution for BNS population [ PDF ].
    • Isabel's search code [ HTML ]

Contact Info

InterCall telecon service:
   Phone: 1-866-380-5536, participant code: 435 672 9587 #
   International callers ++1-816-249-4731 with same code
   International access numbers (~10 cents/min surcharge for LIGO Lab):
      Paris 017 080 7156 ;
      Lyon 042 603 0036 ;
      Germany toll-free 0800 182 1591 ;
      Berlin 030 726 167 371 ;
      Rome 00645 217 080 ;
      UK toll-free 0808 234 7914 ;
      London 0203 107 0293
   Handy participant command:  Press *6 ["*M"] to mute, #6 to unmute


  1. Continued review of GRB070201 search
    • Isabel has checked the agreement between the theoretical non-central beta distribution and the observed distribution in her Monte Carlo. These are seen in the following plots:
      • 25 ms cross-correlation [ PNG ]
      • 100 ms cross-correlation [ PNG ]
      This shows that the distribution is accurately modelled by the non-central beta function and that the value of lambda90% is calculated correctly.

      However, it is still unresolved whether the hrss90% upper limits are correct. The central question hinges on whether Eq. (38) of Jolien's notes correctly relates hrss and lambda.

      Isabel is concerned that the approximation to obtain Eq. (29) is invalid. It is certainly not true for lambda90%. But this equation is only used to relate rho and lambda in Eq. (35), which can be done in the low SNR limit since all that is required is the proporionality constant.

      One possibility is that hrss is not computed to include both the plus and the cross polarizations -- just the injected polarization. If this is the case then the value of hrss that Isabel needs to get the required value of lambda90 would be a factor of sqrt(2) too high. This could bring the result in agreement with the expectations. Isabel has done a quick check of this and thinks hrss is being correctly computed.

      Action: Jolien needs to check the relationship between lambda and hrss in his notes.

      Action: need to check the calculation of hrss for circularly polarized sine gaussians. The easiest way to do this is to compare the strain timeseries that Isabel obtains for a fixed hrss with one of the MDC frame contents for the same value of hrss. Also need to identify the piece of code that computes hrss. It should be somewhere in this file [ HTML ].

    • The issue of the inspiral sensitivity has not been addressed yet. Various possibilities have been discussed (including the 30% calibration error issue). Need to firm up exactly what has been done.

      Isabel says that the largest cross-correlation occurs when the inspiral signal is at 130 Hz.

      Action: Isabel will send Jolien the cross-correlation values as a function of time that went into the plot (currently Fig. 7 in the draft paper) so that the exact time of the largest cc value can be found. Jolien will try to find the true value of hrss,det for this injection and compare to the expected value of the largest cc.

    • Peter Kalmus has agreed to update his code review of the crosscorrelation code.

      Action: Isabel will put the most recent version of the code into CVS so that the diffs are available. She will contact Peter when this is done.

      Action: Peter will go through the changes in the code and update his code review. Of particular importance is checking the calculation of hrss in the case of circularly polarized sine-Gaussian signals (see above), and the handling of injections of MDC frame signals.

    • Paper is currently in flux. Reviewers need not spend time preparing comments on it yet. Expect a more stable version on Wednesday.
    • One issue of presentation arises. Our understanding is that the inspiral group is considering two primary scenarios: (i) there is an inspiral in M31 -- what is the probability that such a signal could be missed (e.g., due to poor orientation); (ii) supposing that there is a optimally-oriented inspiral associated with the GRB in the direction of M31, what is the minimum distance that the source could be (90% confidence).

      The primary burst contributions should be a statement similar to (i) but resulting from the burst search. This would represent a fall-back scenario for the possibility that the phase evolution of the chirp is incorrectly modeled. (Is this really interesting.) A statement similar to (ii) could be done using Monte Carlos or (a quick and dirty version) using the hrss upper limit and the assumption of optimal orientation could yield a distance limit. But the most important result for the burst section would be a energy statement that could be used in interpreting other scenarios such as the possibility that the GRB was actually a SGR burst from a neutron star in M31. For this purpose, the energy that should be quoted is the isotropic energy. Recommend that this should be the focus of the burst section of the paper.

      In summary, the burst section should concentrate on two results: (1) rule out inspiral-like signal in M31 with some confidence as a "backup" to the inspiral search; (2) set some isotropic energy, E_GW^iso, limits on SGR model (or other scenario).

  2. AURIGA-LIGO: No report this week.
$Id: minutes-2007-06-11.html,v 1.2 2007/06/11 20:17:37 jolien Exp $