Gaia (E)DR3 extinction law

In the broad Gaia passbands for Gaia (E)DR3, the extinction coefficients depend both on the source spectral energy distribution and on the extinction itself (e.g., Jordi et al. 2010). Following the method presented in Danielski et al. (2018), we compute the extinction coefficient in the m band k_m = A_m/A₀, with A₀ the extinction at 550 nm, as a function of the star's intrinsic colour or effective temperature (both denoted by X):

k_m = a₁ + a₂X + a₃X² + a₄X³ + a₅A₀ + a₆A₀² + a₇A₀³ + a₈A₀X + a₉A₀X² + a₁₀XA₀²

The above formula was fitted on a grid of extinctions convolving the Gaia (E)DR3 passbands presented in Riello et al. (2020) with Kurucz spectra (Castelli & Kurucz 2003) and the Fitzpatrick et al. (2019) extinction law for solar metallicity, 3500 K < T_eff < 10000 K in steps of 250 K, and 0.01 < A₀ < 20 mag with a step linearly increasing with 0.01 mag. We fitted the formula for main sequence stars assuming log g = 4.5.

We present another fit adapted for the top of the HR diagram (that is for giants and the top of the main sequence, up to M_G ∼ 5 mag) where we adapt the surface gravity of the Kurucz spectra to the temperature with log g = 4 for T_eff > 5250 and log g = −8.3 + 0.0023 T_eff for cooler stars.

We performed both fits for various flavours of X = (G_BP− G_RP)₀, (G − K)₀, and TeffNorm = T_eff / 5040 K and for the bands m = G, G_BP, G_RP, J, H, and K, with the near-infrared bands (J, H, K) corresponding to the 2MASS ones. The fit itself has a maximum uncertainty of 3.5%, 1.5%, and 1% in the G, G_BP, and G_RP bands, respectively. However, the main uncertainties are linked to the systematics (choice of the extinction law and of the reference spectra). The coefficients should not be extrapolated outside the extinction and temperature ranges used, the latter corresponding to -0.06 < (G_BP − G_RP )₀ < 2.5 mag or −0.1 < (G − K)₀ < 4.0 mag. The resulting coefficients are provided here in a zip file for download.

For G_RVS, the passband is narrow enough so that a single value of k_GRVS = 0.5385 (value at the central wavelength of the G_RVS filter) can be used, it does not deviate by more than 0.2% in the temperature and extinction range covered above. More information is contained in the paper Gaia Data Release 3: Estimation of the magnitude G_RVS using the RVS spectra by Sartoretti, et al., 2022 and from the Gaia (E)DR3 passbands page.

A Python tool to convert the monochromatic extinction A0 to specified passbands has been made available by the Gaia Data Processing and Analysis Consortium. More details can be found on the Gaia (E)DR3 extinction coefficients in various passbands page.

References:

• Castelli F., Kurucz R.L., 2003, In: Piskunov N., Weiss W.W., Gray D.F. (eds.) Modelling of Stellar Atmospheres, vol. 210 of IAU Symposium, A20
• Danielski C., Babusiaux C., Ruiz-Dern L., Sartoretti P., Arenou F., Jun. 2018, A&A, 614, A19
• Fitzpatrick E.L., Massa D., Gordon K.D., Bohlin R., Clayton G.C., Dec. 2019, ApJ, 886, 108
• Jordi C., Gebran M., Carrasco J.M., et al., Nov. 2010, A&A, 523, A48
• Riello M., De Angeli F., Evans D.W., et al., Dec. 2020, A&A, 649, A3

When using the auxiliary data in these files, we ask you to recognise the ESA/Gaia/DPAC teams by adding an acknowledgement to your work as follows: "This research or product makes use of public auxiliary data provided by ESA/Gaia/DPAC/CU5 and prepared by Carine Babusiaux”.