Yesterday, over 16,000 new asteroids reported by Rubin received provisional designations from the MPC. These objects appear to have bypassed the usual MPC process for survey reports—where objects are first sent to the ITF, then identified as unique through orbital fitting before being assigned a provisional designation. Instead, it seems the MPC checked the orbits directly upon reporting and assigned provisional designations without going through the ITF.
Here’s the question: if all detected moving objects are being reported, then naturally there should be tracklets stored in the ITF without provisional designations. Objects detected only during a single night fall into this category. It is impossible for every object to be detected without fail over three or four nights. However, in reality, even in yesterday’s reports, and further back in the reports from the first look in June last year, Rubin observations sent to the ITF appear to be almost nonexistent.
Does this mean that Rubin reports only include objects identified (via HelioLinC, etc.) as the same object detected over multiple nights, out of all the moving objects detected?
If so, do you plan to report the detection of the single night that could not be identified with other tracklets until the end as an observation sent to the ITF to the MPC?
To add a few details: we do intend eventually to submit un-linked tracklets with some minimum number of sources – probably three or four. We’re still working out the details of this because it isn’t expected to be a major source of discoveries for LSST, for reasons I’ll explain below.
The LSST survey cadence dictates that most sky areas will be observed only twice per night, meaning that tracklets will consist of just two sources. In other words, the LSST will produce mainly ‘two-point tracklets’. This is a major departure from other surveys, which typically aim for four-point tracklets (or three-point tracklets at the very least). A four-point tracklet indicating ‘straight line’ motion on the sky at a constant velocity is a very good indication of real asteroid. By contrast, many two-point tracklets are random associations of unrelated things (different objects or noise detections of various kinds). For this reason, two-point tracklets cannot be submitted to the MPC without overwhelming the ITF (and the NEOCP) with spurious data. The best way to prove unambiguously that a given two-point tracklet does correspond to a real asteroid is to link it to other tracklets on other nights. That’s why the LSST discovery paradigm relies on multi-night linking rather than tracklet submissions: we need the multi-night linkages to know which tracklets correspond to real asteroids.
That said, we will sometimes observe a field more than twice per night, and a small minority of LSST-produced tracklets will have more than two points. These we intend to submit to the ITF – or the NEOCP if the object’s angular velocity is characteristic of a near-Earth asteroid. Many-point tracklets with NEO-like angular velocities will be submitted promptly. Many-point tracklets with angular velocities characteristic of main belt asteroids or more distant objects may not be submitted right away (we’re still deciding on the timescale) because it would be more efficient to incorporate them into a multi-night linkage prior to submission, if this is possible. In any case, all plausible many-point tracklets that do not form part of a multi-night linkage will eventually be sent to the ITF.
Thank you for the very, very clear explanation.!
I had completely assumed that since the Rubin survey began, numerous NEO candidates with only two points would be sent to the NEOCP every night, but that’s not actually the case. Does that mean the expectation that Rubin would monopolize many NEOs in the southern hemisphere isn’t necessarily correct?
And does the search range for links to two-point tracklets within Rubin only include objects observed by Rubin itself? For example, if an object is immediately identified as known, would it be reported even as a two-point tracklet? Also, if a link can be established with an object from another station stored in the ITF, would it be reported as a two-point tracklet or a two-point tracklet over two nights, and then the identification submitted?
For example, DECam has reported numerous MBAs of ~23 mag and Subaru has reported numerous MBAs of ~26 mag to ITF. Therefore, with just two nights of Rubin two-point tracklet data, it can link these to past observations and submit a provisional designation for a new object.