8.1 SSO data in Gaia DR3
Author(s): Paolo Tanga and Alberto Cellino
Gaia DR3 includes Solar System object (SSO) observations collected during the nominal operations of Gaia, from 5 August 2014 to 28 May 2017.
In this chapter, where epoch data are concerned, we adopt the following nomenclature: a transit refers to a transit of a source on the focal plane; an observation is the measurement obtained from the signal of a single CCD. Note that the observation applies to astrometry only, whereas the photometric G-band data are provided at transit level.
SSOs include objects from different classes. Following the nomenclature officially adopted by the International Astronomical Union (IAU) at the General Assembly in 2006, the classes are those of planets, dwarf-planets, satellites, and small Solar System bodies.
In the case of Gaia observations, an a priori choice of targets was introduced, according to the criteria explained in Section 8.2.1. The final number of objects and observations present in the data release is smaller than the number of objects in the pre-selected list. This is due to several reasons, such as the lack of adequate calibration for some observations and the selection introduced in the different processing steps, outlined in the following sections, to reject outliers.
The final list of SSOs present in Gaia DR3 includes small Solar System bodies and planetary satellites. As shown in Section 8.2.1, the vast majority of small Solar System bodies are asteroids, but transneptunian objects (TNOs) have also been selected. Comets are not present.
Concerning asteroids, we distinguish two categories: objects that are in the numbered category of the Minor Planet Center, with good orbital accuracy; and, objects that are identified as moving but are not matched against a numbered asteroid. For objects in the latter category, bundles of observations are provided, corresponding to sets of subsequent transits over a few hours’ or days’ time (Section 8.2.1).
The Gaia DR3 SSO data set contains a total number of observations, the measurements recorded by the single astrometric CCD detectors, equal to 23 336 467, corresponding to the transits of 158 152 objects across the Gaia focal plane. As detailed in Section 8.2.1, a fraction of transit bundles, whose positions do not match those of known asteroids (unmatched asteroids), have been found in the data using a specific procedure.
Both unmatched asteroids and satellites in the archive table have the field number_mp set to zero. Transit bundles that are not matched against known asteroids are given a unique identifier in the field denomination. The identifier is composed of the string Gaia-DR3SSO- followed by a number of 8 digits (for instance, Gaia-DR3-SSO-34117155).
For satellites, the field denomination contains their IAU standard name, plus the identifier composed of the first letter of the planet and the ordinal position in roman numerals. For instance, in the case of Thetys, the third satellite of Saturn, the field contains Thetys (SIII).
For the first time, Gaia DR3 contains a large sample of asteroid reflectance spectra (60 518 objects). per cent of the transits have an associated G-band photometric flux.
The general distribution of the observations over the time period is rather homogeneous, with exceptional gaps, usually shorter than a few hours, due to maintenance operations, such as station keeping manoeuvres, telescope refocusing, micro-meteoroid hits, and other events.
At shorter time scales, detections are clustered in time along a general pattern related to the rotation of the satellite (a period of 6 hours) and dominated by the ecliptic crossing by the two fields of view, at intervals of 106 minutes (from the Preceding Field of View PFOV to the Following Field of View FFOV) and 254 minutes (from FFOV to PFOV). The peaks are strongly modulated in amplitude by the evolving geometry of the scanning plane with respect to the ecliptic. For a general introduction to Gaia operations, see Gaia Collaboration et al. (2016b). During the first part of the period covered by the observations until 21 August 2014, a special scanning mode was adopted to obtain a dense coverage of the ecliptic poles (the Ecliptic Pole Scanning Law or EPSL). In this peculiar geometry, the scanning plane was nearly perpendicular to the ecliptic, with a gradual drift of the longitude of the node at the rate of the Earth revolution.
A smooth transition then occurred towards the Nominal Scanning Law (NSL) that has been maintained constant for the rest of the mission. In this configuration, the spin axis of Gaia precesses on a cone having the axis in the direction of the Sun and an aperture of 45, over a period of 63 days. As a result, the scanning plane varies in inclination with respect to the ecliptic between 90 and 45, and its nodal direction has a solar elongation between 45 and 135.
By considering the possible orientations of the intersection of the scanning plane with the ecliptic, as shown in Figure 8.1, it becomes clear that asteroids are never observed at less than 45 elongation from the Sun and from the anti–solar direction (opposition).
The specific geometry of observations has important consequences for SSO observations. In fact, not only are main-belt asteroids (MBAs) observed at non–negligible phase angles (the Sun-object-Gaia angle), but also in a variety of configurations (e.g., high and low proper motion, smaller or larger distance). The geometry influences many scientific applications and can affect the detection capabilities of Gaia and the accuracy of the measurements.
As extensively explained in Gaia Collaboration et al. (2016b), the satellite rotates at a constant rate, and the images of the sources on the focal plane drift continuously (in the Along Scan direction, AL) across several CCD strips. A total of 9 CCD strips are crossed in the Astrometric Field (AF, numbered from 1 to 9). All SSO astrometric and photometric measurements published in Gaia DR3 (as, previously, in Gaia DR2) are solely based on AF measurements. Each CCD tracks the motion of the image by a Time Delay Integration (TDI) mode, at a rate corresponding to the drift induced by the satellite rotation. The TDI rate is calibrated on the stars and the exposure time is determined by the crossing time of a single CCD, namely 4.4 s. Lower exposure times are obtained when needed to avoid saturation, by intermediate read-out registers (the so-called gates) that are activated at different thresholds when bright sources are detected.
To drastically reduce the data volume processed on board and transmitted to the ground, only small patches around each source (windows) are read-out from the CCDs. For the vast majority of the detected sources (apparent magnitude mag), the window has a size of pixels, but the pixel value is integrated in the Across Scan (AC) direction. Only one-dimensional (1D) information, in the AL direction, is thus available, with the exception of the brightest sources ( mag) for which a full two-dimensional (2D) window is transmitted. Sources of intermediate brightness are given a slightly larger window () but AC binning is always present.
As the TDI rate corresponds to the nominal drift motion of stars on the focal plane, the image of an SSO, due to its apparent motion with respect to the stars, generally exhibits trailing in the direction of its motion. Moreover, it tends to move away from the window centre during the transit, with a more or less relevant flux loss. For instance, the signal of an asteroid with apparent motion (in the AL direction) of 13.6 , will move a distance of one pixel during a single CCD crossing, with a corresponding image smearing. The central peak of its signal will reach the edge of a 12–pixel–wide window while travelling from AF1 to AF5. In practice, the accurate determination of the position of the source within the window will be affected by an increasingly larger uncertainty over the transit, and become unusable from AF5 (or AF6) to AF9. This is thoroughly explained in Section 8.3.3.
The orientation of the scanning plane of Gaia is most of the time not far from being perpendicular to the ecliptic plane. In this specific orientation, asteroids having low–inclination orbits move mainly in AC direction. As the AC pixel size is 3 times larger than AL, the effects of motion are strongly reduced. However, during the precession cycle the scan plane reaches inclinations down to 45 with respect to the ecliptic. These variations in orientation, coupled with the distribution of the inclinations of asteroid orbits, translate into a large range of possible orientations of the motion vector on the (AC, AL) plane. For any object, each transit on the focal plane, occurring at a different epoch and position on the sky, is a unique event, associated with a non-repeatable relative geometry of motion and scanning direction.
The apparent displacement of SSOs at the epoch of each observation is clearly a major factor affecting the performance of Gaia at each transit. Other effects acting on single CCD observations exist, such as local CCD defects, local Point Spread Function (PSF) deviations, cosmic rays, and background sources. For all these reasons, the exploitation of the single data points must rely on a careful analysis, taking into account both the geometric conditions of the observations and appropriate error models.
A direct consequence of the observation strategy employed by Gaia is the peculiar error distribution for each single astrometric observation. In most cases, due to AC binning, accurate astrometry in the AFs is available only in the AL direction. Of course, this is a natural consequence of the overall mission design, aimed at converting an accurate measurement of time (the epoch when a source image crosses a reference line on the focal plane) into a position. The AC information thus remains approximate due mostly to binning, but also due to attitude uncertainties and to the lower resolution of the telescope in the AC direction. In practical terms, the astrometry is reduced to a single dimension (AL).
As illustrated in Figure 8.2, the resulting uncertainty on the position can be represented by an ellipse extremely stretched in the AC direction. When such position is converted into another coordinate frame, such as Right Ascension (RA) and Declination (DEC) in the equatorial reference frame, a high correlation appears for the related RA and DEC errors. Taking into account such correlation is essential to exploit the full accuracy of Gaia astrometry. The data processing pipeline has been designed in such a way as to take into account this peculiarity, to provide in output the uncertainties and the correlations.
The data processing for SSOs is illustrated by the simplified scheme in Figure 8.3. The pipeline works on the basis of a predefined list of transits of SSOs in the FOVs of Gaia. A list of accurate predictions has been therefore created as a first step, by matching the evolving position of each known object to the sky path of the Gaia FOVs. The predictions of SSO transits have then been matched to the observed transits, provided by the Initial Data Treatment (IDT) output (Gaia Collaboration et al. 2016b).
The core of the pipeline collects all the data needed to process the identified transits (epoch of transit on each CCD, flux, AC window coordinates, and other auxiliary information). A first part, labelled ‘Identification’ in the scheme, is in charge of computing auxiliary data for each SSO, and to assign a correct label identifying the object. Focal plane coordinates are then converted to the sky reference frame using the Astrometric Global Iterative Solution (AGIS) and the corresponding calibrations in the astrometric reduction described in Section 8.3.3.
Two separate procedures are responsible for deriving transit-level G-band photometry and mean reflectance spectra for asteroids. They are described in Section 8.3.5 and Section 8.3.7, respectively. Moreover, some specific validation procedures perform the task of identifying and rejecting anomalous data before recording them into the archive. The details of all the developed procedures are explained in the following sections.