Gaps in Gaia eDR3 and DR3 data

A key element for a proper scientific interpretation of Gaia data is knowledge of various selection functions. In turn, an essential part in establishing reliable selection functions is knowledge of the gaps that exist in the data, for instance those caused by spacecraft events such as refocussing, decontamination, or anomalies.

This webpage provides tables with gaps in the data streams and data products of Gaia's third data release (Gaia EDR3 and Gaia DR3). Since astrometric, photometric, and spectroscopic data have been treated in different pipelines, separate gaps tables exist for each data type. The below tables exclusively apply to Gaia EDR3 and DR3; separate tables will be published for future data releases. Gaps in Gaia DR2 can be found here.

The definition of gaps applied here is somewhat arbitrary and, moreover, erroneously suggests that a black-and-white situation exists. For instance:

• As a result of on-board processing limitations, Gaia will produce incomplete coverage for BP/RP and RVS spectra since only a fraction of the transits can have a window allocated by the Video Processing Unit (VPU). This is generally a function of magnitude but also of sky position (crowding). The effect can be very severe, reaching more than 30% loss in coverage for faint objects in dense regions.
• Gaps in the published data are also caused by the occasional on-board deletion of science data in times of on-board storage overflow. Such gaps can show magnitude-dependent effects due to the priority management of the science data on board. These gaps are generally not covered here except for a few major outages.
• "Bad" CCD observations have been downweighted in the Gaia EDR3 astrometric solution through a statistical weight factor that depends on the size of the post-fit astrometric residual in relation to the expected uncertainty – see Equation (66) in Lindegren et al. (2012). If such downweighting is systematic over certain periods of time (and evidence exists that such situations do exist), it can also create gaps.
• A number of time intervals exist in which only six of the seven VPUs were generating science data, leading to partial gaps. Such (partial) gaps are not covered here in the astrometric or photometric tables.
• Observational data has been removed from the processing streams due to, for instance, gaps in the offset non-uniformity libraries and/or failures in the offset non-uniformity determination and/or gaps in the background (straylight) correction maps. Moreover, observations obtained with non-nominal configurations (e.g., gated faint-star windows) quite often have poor calibrations with a higher rate of failures when applied.
• Various filters applied during the Gaia EDR3 and DR3 data processing have excluded data, for instance observations that have been flagged as problematic in the acquisition or image-parameter determination processes or outliers originating from cross-match problems. Also, truncated windows have not yet been processed in Gaia EDR3 or DR3 and neither have so-called complex TDI-gate configurations.
• Both data releases have employed "artificial" selection criteria that exclude data (e.g., radial velocities are not yet published for objects fainter than 14^th magnitude or for objects with very cool or very hot cross-correlation template spectra).
• Finally, a complex quality filtering scheme has been applied to select which sources are published in the actual data release (see Fabricius et al. 2021 for the Gaia EDR3 filters).

Astrometry

Section 2.2 of Lindegren et al. (2021) mentions that the Gaia EDR3 astrometric processing has used data spanning revolutions 1192.1295–5230.0880, with Table 1 in that paper providing 26 "gaps longer than 1 revolution (0.25 day) and some other events of relevance for the calibration model". Using internal data sources, we have extended this list and have created a merged table containing 138 unique gaps. Note: micro-meteoroid hits after OBMT revolution 3750.5 (23 May 2016) do no longer cause gaps but are taken into account through a corrective attitude model combined with downweighting of observations.

Photometry

Appendix D of Riello et al. (2021) gives an extensive discussion of gaps, concluding that "Table D.1 provides a list of the [7] known gaps in the Gaia EDR3 photometry … [while] Table D.2 provides the [92] time ranges for which [G-band] epoch observations were excluded from the source photometry [because the calibration could not track the fast changes in system response]”. We have merged these Tables D.1 and D.2, leading to 7 + 92 = 99 gaps that are contained in this table. Note: in two cases — around "Decontamination 4" and "Decontamination 5 (including basic-angle test)" — tiny overlaps exist between adjacent gaps. These overlaps are of duration 0.06 seconds (0.000003 revolutions) and no correction has been made in the table.

It is important to stress that users who are concerned with selection functions also need to consider the effect of the changing magnitude limit of the survey resulting from the continuous build-up of contamination followed by "instantaneous" decontamination events. As can be seen in Figure 5 in Riello et al. (2021), this effect can be larger than 0.5 mag and is different between the three passbands, with BP affected most, followed by G, and RP affected least.

Spectroscopy

In the spectroscopic case, the granularity of the gaps is per CCD row (with each CCD row corresponding to one Video Processing Unit). With this understanding, a list of 2829 gaps has been identified based on three different sources:

1. 177 occurrences (multiplied by 4 since equally applicable to each CCD row, so 708 entries) corresponding to missing astrometry plus associated reconstructed spacecraft attitude. These gaps are denoted by "AGIS" (indicating the absence of an Astrometric Global Iterative Solution for the associated time interval).
2. 2111 gaps corresponding to missing ASD7 object logs. These gaps are denoted by "ObjectLog”.
3. 10 gaps representing tests on the VPU application software version 2.8 (mostly but not exclusively on CCD row 5), which have been added manually. These gaps have been labeled "Manual”. 

The actual list of 177 × 4 + 2111 + 10 = 2829 spectroscopic gaps applicable to Gaia DR3 is available here. Note: numerous overlaps exist between gaps. These have not been removed.

Table description

The description of the contents of each column in each of the data files is as follows:

• Column 1 = start time of the gap, in revolutions as counted in the on-board mission timeline (OBMT). The relation between OBMT and calendar date and time is provided in the Gaia EDR3 on-line documentation.
• Column 2 = end time of the gap, in OBMT revolutions.
• Column 3 = duration of the gap, computed as end time minus start time, in OBMT revolutions. One revolution is 21,600 s.
• Column 4 = short description of the gap.
• Column 5 = N/A (astrometry), instrument (photometry), CCD row (spectroscopy).
• Column 6 = N/A (astrometry), N/A (photometry), origin of the gap (spectroscopy).