Q 1: How many period/sessions do we need to determine the presence of a satelite?
The procedure is following:
Times needed for the three observational phases:
Reduced photometric data should be sent both to Petr Pravec (ppravec@asu.cas.cz) and Peter Kusnirak (peter@asu.cas.cz), unless other arrangement has been made for a particular asteroid.
1. Primary's rotation period is established uniquely and with (at least) a double coverage of all rotational phases.
2. If there is found a significant (>0.04 mag for our Survey) deviation from the single-periodic curve suggesting a possible attenuation due to a mutual event, the asteroid becomes a binary candidate and more observations (from a few stations spread over the world) are planned.
3. When the event is confirmed with more detections on following nights and its (orbital) period is derived uniquely, then it becomes a full, regular detection of a binary asteroid.
1. Primary's rotation period may be doubly covered even on a single long night (if there is no gap in the coverage longer than 0.5 h),
but more typically, it takes 2-3 nights. Some cases (e.g., periods of several hours that cannot be covered fully during shorter nights, or asteroids with more complicated lightcurve shapes) may
require to take a few more nights.
2. + 3. Each mutual event -there are normally two events per orbital period- needs to be covered twice at least, and orbital phases
outside of events need to be covered as well to a degree that the orbital period is derived uniquely and size ratio between the two components of
the binary system estimated from the depth of the (total secondary) event. It typically takes several nights from a few collaborating stations.
Q 4: What is a procedure with points deviating from a single periodicity?
A check of individual (isolated) deviating points is not critical (for other purpose than perhaps just observer's learning and tuning/improving his system); such isolated outliers are supposed to be observational artifacts, marked so in analysis, and no significant conclusion is implied from them. So, it is not necessary for the observer to invest much time in checking them. A check of a series of systematically deviating points is, on the other hand, always needed. After such check (and no observational problem found), we consider such points to be a suspected feature and need to confirm it with more observations.
Q 5: If I re-reduce a session while checking deviating points, do I have to resend the revised dataset to be included in a new analysis?
If the observer finds no problem in the original reduction, there is no point to send a new reduction if it is virtually the same (with only insignificant, statistical variations in reduced point values) as the original reduction.
Q 6: For slow rotators, do we need to get as thorough data as for faster rotators?
For asteroids with periods longer than 10 hours where there is a very low probability to find them being asynchronous binary (as an expected time scale for them residing in the P>10 h domain and still not being fully synchronized is very short), it is needed only to demonstrate that they are longer periods than 10 hours, but a complete unique solution with a full coverage is not required. Some observers want to invest more of their observing time and work even such slow rotators more thoroughly anyway. Though such more thorough coverage is not necessary for the BinAstPhotSurvey and it is not pursued, it is of course no problem for the Survey to process better data even for the slow rotators; such P>10h target is kept open as worked in the Binast-Webtool until the observer decides him/herself that he/she has already got enough data for the slow rotator and calls it over.
Q 7: What criteria are being used to make a discovery announcement on CBET?
The basic criterion on when a binary discovery is to be announced on CBET is following: P_1, P_orb, and D_2/D_1 estimated uniquely, or a lower limit for D_2/D_1 if more data during the apparition are not likely to nail down D_2/D_1.
Occassionally, we consider some additional, "second order" criteria. E.g., where there is seen an obvious secondary variation, then we may want to nail down P_2 as well before announcing the binary on CBET. But such (or occassional, rare other features) are never allowed to postpone the publication by a longer time.
Two supplementary conditions are that a repeated coverage of both events, and a coverage of full orbital period are needed. The two conditions are in place in order we can check assumptions of the model and quality of data.
Last update: 2007 November 20