14.6 Time series and variability

Author(s): Tatiana Muraveva

Gaia DR3 publishes a sample of 11.7 million sources which includes variable stars of about 30 types of variability and about 1.3 million sources (variable or not) in the Gaia Andromeda Photometric Survey (GAPS). For all of them, time-series photometry in the $G$, $G_{\mathrm{BP}}$ and $G_{\mathrm{RP}}$ bands is released. Here, we present a brief overview of some issues found during the scientific validation of this dataset. Interested readers are encouraged to refer to the relative papers for more details.

• •

  A number of sources are classified as variables of more than one type of variability. While most overlaps between different types can be scientifically explained, some stars have contradicting classifications (e.g. 3159 sources classified as long period and short-timescale variables). Detailed analysis of this issue and final classification for these sources are provided in Lebzelter et al. (2023).

• •

  Intensity-averaged magnitudes in the $G_{\mathrm{BP}}$ (int_average_bp) and $G_{\mathrm{RP}}$ (int_average_rp) bands for four and two RR Lyrae stars, respectively, have unreliable negative values reaching $G_{\mathrm{BP}}=-88\pm 22$ mag. These six sources are faint RR Lyrae ($G\sim 18.5-19$ mag) for which the Specific Objects Study pipeline for Cepheids and RR Lyrae (SOS Cep&RRL, Clementini et al. 2023) failed to fit data points with the model line in the $G_{\mathrm{BP}}$ and $G_{\mathrm{RP}}$, hence, providing unreliable values (see Clementini et al. 2023 for further details).

• •

  Absorption in the $G$ band (g_absorption) takes a negative value, but still consistent with zero within the errors for 9503 RR Lyrae. Moreover, for several RR Lyrae, absorption in $G$ reaches unreliably large values up to $A_{\mathrm{G}}=3367\pm 320$ mag, likely caused by the imprecise estimation of the $G_{\mathrm{RP}}$ magnitudes for the faint sources (see Clementini et al. 2023 for further details).

• •

  Double-mode RR Lyrae stars (RRd) are known to pulsate in both, fundamental and first-overtone modes, with a typical ratio between the first-overtone and fundamental periods $P_{1o}/P_f\sim 0.74-0.75$. However, 121 RRd in the Gaia DR3 have a ratio of periods $P_{1o}/P_f\sim 1$. The reason is the re-mapping issue in the SOS Cep&RRL pipeline. Table with correct periods for these 121 RRd stars will be provided in Clementini et al. (2023).

• •

  Uncertainties of the duration [phase fraction] of the primary (derived_primary_ecl_duration_error) and secondary eclipses (derived_secondary_ecl_duration_error) have unreliable values of more than one for 7 and 62 eclipsing binaries, respectively. At the same time, derived_primary_ecl_depth_error, derived_secondary_ecl_depth_error, geom_model_gaussian1_depth_error, geom_model_gaussian2_depth_error take ‘Null’ values for a number of eclipsing binaries, while corresponding parameters are not ‘Null’. More details on this issue can be found in Mowlavi et al. (2023).

• •

  For three RR Lyrae, uncertainties of the intensity-averaged $G_{\mathrm{BP}}$ magnitude (int_average_bp_error), intensity-averaged $G_{\mathrm{RP}}$ magnitude (int_average_rp_error), peak-to-peak $G_{\mathrm{BP}}$ amplitude (peak_to_peak_bp_error) and peak-to-peak $G_{\mathrm{RP}}$ amplitude (peak_to_peak_rp_error) are equal to zero, while the corresponding parameters are not zero. More details on this issue can be found in Clementini et al. (2023).