The undersigned gravitational wave detector groups who are currently operating gravitational wave detectors, agree to set up a data exchange protocol called the International Gravitational Event Collaboration, for coordinating and correlating the data from the detectors. It is based on some technical and policy rules for data exchange and provides the constituition of one or more facilities, open to the members with rules to be established, to maintain a data-base and a news group.
The following sets out the principles of the protocol. The details will be defined and updated according to experience and the changes in computer technology.
1 - Technical points:
- Each group will set up a site (ftp, www by password, telnet, ...) on which data will be continually available in agreed formats and normally with an updating rate close to the real-time and not exceeding one day.
- The data will consist of ”candidate events” and service information such as parameters of the antenna, start run and stop run times, hourly statistics of the noise and descriptions of the apparatus and of the filters used for data processing.
- A news group is to be maintained to allow planning and coordination of detector maintenance and downtime. LSU volunteers to set this up.
- The groups will work towards automatic analysis which will, in the case of very rare coincidence events, produce an automatic mail alert.
2 - Physical points:
- For each operating antenna the individual group will select, with a threshold matched to the noise of the detector of order 100 events per day. The technique to produce the events will be chosen by the individual group on the basis of their experience and the type of noise of their detector.
- The groups are committed to set up procedures to reduce spurious candidate events.
- The groups will describe, post and update the procedures used to reduce spurious candidates.
- The minimum information about each event consists of - the time (in UT) of the event maximum – the amplitude in units of ”standard burst strain” - the duration of the event - the mean noise of the detector at the time of the event in units of h.
- Additional fields, giving further information on candidate events, may be provided at the individual groups discretion. For instance if any group uses more filters on the data, it will be possible to post additional event lists from these filters.
- The time accuracy will be at least 0.1 seconds and the groups will perform procedures to check this and will give information in case of larger errors.
- The exchange of raw or filtered sampled data will be considered in the future.
3 - Policy points:
- The goal of the IGEC is to standardize and simplify the data exchange between the groups and to mantain a continous discussion on the data and the analysis procedures. The members of this center are committed to not make public results based on this data exchange without unanimous agreement of all members.
- The core membership of the center consists of groups producing gravitational wave data. Membership can also be given to groups for data analysis purposes.
- Members will be added or dropped by unanimous agreement of the core members.
- The data of the collaboration will become public after a certain period to be established.
William Hamilton (LSU)
Warren Johnson (LSU)
Guido Pizzella (Rome)
Giovanni V. Pallottino (Rome)
David Blair (UWA)
Mike Tobar (UWA)
Massimo Cerdonio (Padova)
4 July 1997 at the 2nd Amaldi Conference