Hi! I’ve been reviewing the Data Product Definition Document and Early Science plan and have a few clarifying questions about the prompt product database vs what’s in the alerts.
Do all DIASources become alerts? If not, will all the DIASources be in the PPDB even if they’re not associated with an alert?
For the DIAObjects in the table, are there more datapoints than what would have been included in the alert packet?
If an object is in a reference image, is it a DIASource in all epochs even if it’s not changing?
I’m interested in crowded fields. Due to computational expense, would there be more objects in the PPDB than what might be alerted?
Is the PPDB also expected to be “ramped up” like alerts will be?
Yes, all DIASources become alerts and appear in the PPDB.
There may be additional forced photometry measurements in the PPDB for a given DIAObject. These are included in alert packets when available, but depending on the timing there may be forced measurements in the PPDB that have not yet appeared in an alert.
If the object is not changing, science - reference should yield zero flux in the difference image, and no DIASource will be created.
We don’t yet believe that capacity constraints will prevent sending all alerts in crowded fields.
See ls.st/rtn-011 for discussion of the timeline of PPDB availability.
A few followup questions:
2. My understanding is that Rubin sends out one year of forced photometry data with the alert packet. Once Rubin has been going on for longer than a year, will there be more than a year of forced photometry data in the PPDB for a DIAObject if it’s been alerted more than once?
3. This might not have an easy answer, but is there an estimate of what fraction of stars would be noisy enough that they would appear with non-zero flux in the difference image even if there’s no significant variability?