Charter of the LSC Burst Analysis Working Group
LIGO-M060072-03-Z
This draft produced June 15, 2006
Adopted June 23, 2006


MISSION

The Burst Analysis Working Group (or "Burst Group") is a working group
of the LIGO Scientific Collaboration which is devoted to searching for
transient gravitational wave signals without relying on accurate
knowledge of the waveforms.  Because of the wide variety of potential
signals, Burst Group members employ a range of detection strategies,
from focused searches for reasonably well-modeled astrophysical
objects to search methods which are designed to detect signals with
completely arbitrary waveforms.

The Burst Group serves as a forum for communication among people with
related interests, to promote the exchange of ideas and technical
expertise, and to carefully examine analysis methods and results
before they are presented to the full LSC.  Burst Group members should
have the freedom to be creative and to pursue promising scientific
goals, while following good organizational principles and being
receptive to constructive criticism and suggestions from other Burst
Group members.  The ultimate mission of the group is to produce
thorough, excellent scientific results that the LSC can be proud of.


PARTICIPATION

All LSC members are welcome to participate in the Burst Group to the
extent they are interested and able.  The first step is to subscribe
to the group's mailing list in order to receive organizational
announcements and to follow/participate in email discussions.  The
group also collaborates through an electronic notebook (e-notebook)
system, weekly teleconferences, and occasional face-to-face meetings,
all of which are open to all LSC members.

All LSC members who are doing work related to burst data analysis are
expected to participate in the activities of the Burst Group in order
to share that work with their colleagues.  The Burst Group e-notebook
should be used regularly to document progress and results, either in
the form of detailed e-notebook entries or else to provide an overview
and portal to material found on other web servers.  In addition to the
e-notebook, Burst Group members are expected to communicate their work
to the other members of the Burst Group through email, telecons,
and/or meetings, as appropriate.  This communication is essential to
provide exposure, understanding, and feedback for the ongoing work.
The Burst Group Co-Chairs will help schedule presentations, but group
members should be pro-active about finding opportunities to share the
work they are doing.  All active members of the group are expected to
contribute to the data analysis white paper as it is periodically
updated, and to other work summaries (e.g. presentations to the LSC or
to review panels) when requested.  The Burst Group will also
periodically report to the MOU Review Committee progress on activities
covered by LSC group MOUs, as part of the Burst Group's contribution
to the MOU review process.

Many interesting astrophysical analyses can be pursued with the data
now being collected, and the success and reputation of the LSC will
depend on the quality and breadth of completed analyses.  Burst Group
members are encouraged to get involved in astrophysical analyses and
carry them through to completion.  Because these analyses are
undertaken on behalf of the whole LSC, they must follow a set of rules
for coordination, reporting, and oversight purposes.  In particular:

1. Each analysis must be identifiable and understood by the Burst
   Group members.  One or more persons must contact the Co-Chairs to
   schedule an initial presentation to the Burst Group to describe the
   scientific goals and the planned approach.  This provides an
   opportunity for feedback on the goals, methods, and relationship to
   other ongoing analyses.  Scientifically sound analysis plans will
   naturally win the approval of the Burst Group, which should be
   obtained before any substantial work is done on the analysis.  An
   active analysis task should have a distinctive name.  The Burst
   Co-Chairs will maintain a web page with a complete list of active
   analysis tasks.

2. An analysis, even if proposed by a single person, must be open for
   participation by other LSC members who are willing to make a
   commitment to it.  In fact, it is encouraged that any analysis have
   at least two people associated with it.  This prevents the work
   from being carried out in a vacuum and puts an emphasis on regular
   communication, which is ultimately necessary to bring the completed
   analysis forward to the LSC.  Even if one person is doing all the
   real work, a second person can make an important contribution by
   just following along, asking questions, providing suggestions, etc.
   In other cases, a subgroup may naturally form consisting of several
   Burst Group members who work together closely on related analysis
   projects.

3. Ultimately, the entire LSC has to understand and stand behind an
   analysis which is carried out by a small core team.  Good
   communication and documentation is essential.  Every analysis team
   must maintain a "living" (frequently-updated) project summary with
   the goals of the project, names of the people involved, milestones
   reached, projected timetime for future progress, etc.  In addition,
   technical documentation (possibly in the same document as the
   project summary) must be generated as the analysis proceeds in
   order to document the methods used, how they were applied, what
   choices were made and how, etc.  All software (including scripts)
   must be stored in a repository that can be accessed by other
   people.

4. The project summary, technical documentation, and other supporting
   material must be accessible through the e-notebook.  Progress
   reports must be made to the Burst Group at least once every three
   weeks, preferably in a regular Burst Group telecon or meeting,
   otherwise as an e-notebook entry.  Important findings (especially
   detection candidates!) should be shared with the Burst Group
   promptly.

5. An analysis is not complete until it is understood and implicitly
   approved by the Burst Group, and reviewed in accordance with LSC
   standards.  An analysis typically has several aspects which are
   scrutinized in different ways:

   a. The concept of the analysis is effectively reviewed by the Burst
      Group at the time of the initial proposal, and in the context of
      regular progress reports to the Burst Group.

   b. The implementation details of the analysis, including the
      software used to carry it out, typically will be checked by two
      "internal reviewers" drawn from the Burst Group and the Burst
      Analysis Review Committee.

   c. Analysis results will be formally reviewed by the Burst Analysis
      Review Committee.  The technical documentation for the analysis
      must be complete for this review to take place.

   d. Paper drafts will be reviewed by the Burst Analysis Review
      Committee, who must make a recommendation to the LSC and/or the
      LSC Executive Committee before a paper may be posted as a public
      preprint or submitted to a journal.

6. Burst Group members, like all LSC members, must adhere to LSC
   policies regarding presentations and the release of astrophysical
   results.


LEADERSHIP

The Burst Group is led by two Co-Chairs who, between them, should have
significant experience with both instrument science and data analysis.
The Co-Chairs work together as equals to schedule and lead telecons
and meetings, monitor and coordinate burst analysis activities, serve
on the LSC Analysis Committee, and represent the Burst Group in
dealings with LSC standing committees and the LIGO Directorate.  In
the internal workings of the Burst Group, the Co-Chairs mediate
discussion to help the group make decisions by consensus whenever
possible, but may also call for a majority vote or make a (joint)
executive decision if it is necessary to make a decision on a matter
for which there is no clear consensus.

The selection procedures and terms of the Co-Chairs are governed by
the LSC Bylaws and the policies of the Elections and Membership
Committee.  As specified in the LSC Bylaws, the Co-Chairs may either
be elected or be appointed by the Spokesperson; a simple majority vote
of active Burst Group members, polled whenever a Co-Chair needs to be
chosen, will determine which method is to be used.  If this vote ends
in a tie, then the election method will be used.


MODIFICATIONS TO THE CHARTER

Any Burst Group member may propose a modification to this charter by
contacting the Co-Chairs.  The Co-Chairs will review the proposal for
well-formedness (whether or not they agree with the proposal) and will
arrange with the proposer to circulate the proposed change and the
rationale behind it to the Burst Group mailing list.  After allowing
at least one week for discussion in email and in a regular telecon,
the Co-Chairs will call for an email vote on the proposal.  A
two-thirds majority is required for the proposed modification to be
adopted.


HISTORY OF CHARTER

The LSC Bylaws, approved at the March 2006 LSC Meeting (LIGO document
M050172-01), formalized the LSC working groups and specified that each
working group should establish a charter.  The initial version of the
Burst Group charter was drafted in May/June 2006 and was approved by
the Burst Group on June 23, 2006.