<I> SNfactory</I>

SNfactory follow-up pipeline

The output of the search pipeline at LBL will be a list of candidates with various scores indicating any potential problems.
Candidates (perhaps 10) will be inspected and rated by a human supervisor. This will occur at LBL, but all collaboration sites should develop experience in this area.
Based on the final ratings, an observing scheme for follow-up telescopes will be composed and sent from LBL. The details of the observing scheme will depend strongly on the scheduling constraints at the telescopes and the amount of control we have at that telescope. Where we have the most control, the observing scheme will be composed of self-contained scripts that obtain all the observations and calibration for each target. For telescopes with less control (e.g. YALO) a request for observations will be issued.
At the follow-up telescopes a night assistant will execute the observing script or observing request.
The script will send the resulting data to the collaboration sites. Since the amount of follow-up data taken each night should be small (of order 500 MB), several sites should receive the raw data (perhaps from each other rather than the observatory) and maintain database and analysis tools.
There will be a master database which keeps track of what observations have been taken. This will be consulted by the observing scheme script running at LBL to plan subsequent observations.
All follow-up will be reduced automatically (provided the observing scheme was executed successfully and completely). Partial observations (hopefully rare) may need to be treated with some interaction. (The requirements here are very extensive, and will need to be given in more detail elsewhere.)
Once reduced, the results can proceed to the analysis stage. For newly discovered SNe, the first analysis stage will be to classify the SN and to rank it relative to other SNe being monitored. There may be cases when the observing queue is too full to accept new SNe, or in which follow-up for a SN has faltered and that SN is therefore abandoned in favor of a more interesting candiate.
The initial results of the analysis will be sent to the master database, and used to update the observing plan for the next night. For screening observations, the SN redshift, type and lightcurve phase must be determined. In addition information on intrinsic luminosity, metallicity, host galaxy properties is desirable so that SN parameter-space is well explored by the experiment.
Initial results also will be used to determine preliminary lightcurves and spectral series. These will be monitored automatically and by humans to make sure observations are on track (again, with feedback to the observing scheme planner) and to issue alerts of unusual behavior.
Since each site will have follow-up data, further analysis and research can proceed according to the collaboration plan then in place, with each science group undertaking whatever additional work is required for publication.