This catalogue contains photometric stellar parameters for 9+ million stars in common between the SkyMapper survey and Gaia DR2. Columns are SkyMapper 'object_id', Gaia 'source_id', effective temperature 'Teff', uncertainty 'e_Teff', metallicity 'feh' and uncertainty 'e_feh'. Stellar effective temperatures and metallicities have been derived from SkyMapper+2MASS photometry as explained in Casagrande et al. (2019, MNRAS, 482, 2770). Briefly: all photometry has been corrected for reddening as explained in Section 5 of the paper. SkyMapper v band photometry has also been corrected to account for variations of zero-points across the sky, see Eq (5). Effective temperatures come from Eq (10), which is calibrated upon a sample of ~86,000 stars with effective temperatures from the InfraRed Flux Method. Metallicities are derived using Eq (12), which is obtained using Principal Component Analysis on a sample of ~70,000 stars with spectroscopic [Fe/H] from the GALAH DR2 survey.
Comparison to various spectroscopic surveys (see Fig 11 and 13 in the paper) indicates that these photometric [Fe/H] are generally good to within 0.2 dex (rms). The uncertainties quoted in this catalogue are obtained propagatig nominal photometric errors through the Teff and [Fe/H] calibrations. In addition to that, for the effective temperature calibration, a scatter of 33K (the rms of Eq 10) is added in quadrature. A constant systematic of 20K is further included (the uncertainty on the zero-point of the Teff scale). For the metallicity calibration, an rms of 15 dex is added in quadrature, so that the median uncertainty for the sample peaks approximately at 0.2 dex (the precision suggested when comparing to spectroscopic surveys).
The catalogue provided here is restricted to stars passing a number of quality cuts: 2MASS Ks quality flag 'ph_qual' set to either 'A' or 'B', 'gal_contam' = 0, 'mp_flg' = 0, SkyMapper 'class_star' > 0.8, 'prox' > 15, 'v_flags' = 0, 'e_v_psf' <= 0.1, 'g_flags' = 0, 'e_g_psf' <= 0.04, Galactic latitudes > |5| degrees from the plane. This last requirement avoids regions with very high reddening, as well as issues with SkyMapper photometric zero-points and crowding towards the Galactic plane. Further, only stars lying within the range of applicability of the metallicity calibration have been considered, see Eq. (13) and (14). These fiducials limit the lowest metallicity to -2.5 dex. Comparison with synthetic photometry indicates that metallicities of turn-off stars are overestimated for (true) [Fe/H] < -1.5. Also, because of the increasing fraction of carbon-enhanced stars below −2, there can be some stars that in fact are more metal poor than predicted by the calibration (see discussion in Section 5.1.3). On the metal-rich side, beware of stars above 0.5 dex.
If you have any question or you notice anything anomalous with the data, you are very welcome to get in contact with Luca Casagrande (luca.casagrande@anu.edu.au).