I got a question from Julio Camargo that seemed worthwhile to share here:
Dear L. Jones,I’m Julio Camargo (LSST - brazilian participation group).
I was at the LSST Community Workshop in Tucson and there are few points
about data access not clear to me. More specifically,
I’m talking about the access to positions of small solar system bodies
(those of already known objects and of new discoveries).
Will the positions of the known objects be only available from the MPC?
Is it possible to have an early access to this data from the LSST (that
is, before it goes to the MPC)?Will the same hold to the new discoveries?
Thanks in advance for your help.
Best regards,
Julio
So here’s my response! (FWIW, @mjuric and @leanne had no objections! so you could consider this answer a reasonable statement of our current understanding 
)
Hi Julio,You will get more authoritative answers to these questions from Leanne and Mario (cc’d here)
But here’s my understanding:
Known objects will be identified as such (where possible) in alerts – this means that if a known object is seen and measured in a difference image, and can be identified as a known object, then a general ‘announcement’ of the measurement will go out that contains (at least):
time, ra, dec, trail length and position angle, known object identification, and a postage stamp image.These alerts are sent out publicly, generally within 60s of the shutter closing on an image. This means you have access to the astrometry of known objects within 60s of the image being complete, through the alert stream. Note that we may or may not send this information to the MPC, depending on the MPC policy at that time (because these are single-image detections, not tracklets or multiple detections within a night). The point of identifying the objects in the alerts is partly to enable followup of the moving objects, but also partly to enable people interested in other objects to avoid following up moving objects.
For new discoveries, things are a little different: for a brand new, unknown object, detections within a night are linked into tracklets – these may be sent to the MPC, but if they are not linked to a known object, I don’t think the MPC will be releasing these widely, and neither will LSST (there is a lot of junk in here … it’s a resource for linking to other unknown/new objects with observations from other nights, but not necessarily useful as a stand-alone piece of data). If the same object is observed on two more separate nights (and tracklets are formed on those nights also), then we can link the tracklets together into a track and then try to fit an orbit: if an orbit fits, this becomes a ‘new solar system object’ at that time.
At that point, the new object is reported to the MPC, and so astrometry would be available from the MPC, and it will also be added to the LSST Daily Orbit Catalog, which is accessible by anyone with data access rights to LSST. The idea is that reporting these new objects to the MPC and adding them to the LSST catalog would happen within a day or even a few hours of each other; depending on the details of how the prompt data processing plan are hammered out, it may be that we report to the MPC, wait for the MPC catalog update, and then add to the LSST daily orbit catalog, or we may add to the LSST daily orbit catalog immediately and then update the orbit when we get the feedback from the MPC; in either case, this synchronization should typically happen within 24 hours.
LSST will only store the astrometry obtained by LSST; MPC will store astrometry from LSST as well as other telescopes.
Does that answer what you were wondering about?
Lynne
So there you go - access astrometry via alerts (known objects, although with some possibility of mismatches or misses), LSST daily orbit database (this should be very reliable), or from the MPC (which will be even more complete than LSST alone, but may have variable astrometric errors).