Published February 6, 2019 | Version Version 1.0
Dataset Open

Distances for the Gaia RV set with corrected parallaxes

  • 1. University of Oxford
  • 2. University of Lund
  • 3. University of Geneva

Description

The datasets are described in https://arxiv.org/abs/1902.02355 , which is submitted to MNRAS.

Each file contains a distance derivation under different assumptions for the Gaia RV dataset.

Please note the following: to allow for freedom in choice of cuts, we have limited ourselves to just minimal cuts: parallax/parallaxerr > 3, detected G,G_BP,G_RP magnitudes > 0 mag, vlos_err < 10 km/s, n_vis > 5, reasonably measured RV (<5550 km/s), distance and 1/parallax < 10 kpc

Note that these cuts are NOT sufficient to ensure quality of the sample. We strongly advise to use the quality criteria laid out in Schoenrich, McMillan and Eyer (2019), and test for the effects of a cut in galactic latitude b > 10 degrees for most applications.

The file format is .csv (comma separated) with the following columns (a key is in row 1)
index: running number in our sample
sourceid: Gaia Source ID
E_dist: expectation value for the distance, unit is kpc
distm2: second moment of distance probability distribution (use the usual formula to get the variance/dispersion)
distm3: third moment of distance probability distribution
distm4: fourth moment """"
parallax: The Gaia parallax _including_ the offset correction chosen, unit is mas
parallax/parallaxerr : ratio of parallax to parallax uncertainty. Again, both parallax and parallaxerror contain the corrections. Use this and the last column to obtain the parallax uncertainty for quality cuts
x: x coordinate (radial coordinate in local cartesian frame), unit is kpc
y: y coordinate in the azimuthal direction, unit is kpc
z: z coordinate (perpendicular to the plane), relative to the Sun, unit is kpc. Please add the solar offset from the plane that you favor (0.02 kpc, Joshi et al., or the Gerhard & Bland-Hawthorn review are good values)
R: galactocentric radius (in-plane, cylindrical coordinates), unit is kpc
U_g: cylindrical radial velocity in a galactocentric frame (positive taken inwards), unit is km/s
v_phi: azimuthal velocity (V_g, azimuthal direction), unit is km/s
v_z: velocity component upwards perpendicular to the plane, unit is km/s
vlos: line-of-sight velocity as provided in the Gaia RV datasets
vloserr: line-of-sight velocity error as provided in the Gaia RV datasets
gl: Galactic longitude l (degrees)
gb: Galactic latitude b (degrees)
pml: proper motion in l (mas/yr)
pmb: proper motion in b (mas/yr)
pml_err: proper motion error/uncertainty in l (mas/yr)
pmb_err: proper motion error in b (mas/yr)
E_bprp: BP-RP excess flux factor (provided in Gaia catalogue)
excessnoise: astrometric excess noise
Gmag: Gaia G magnitude
Rmag: Gaia G_RP magnitude
B-Rcolour: G_BP - G_RP  (colour in mag)
nvis: number of visibility periods
eclat: ecliptic latitude (degrees)

Each file contains a different set of assumptions that result in a near-zero average distance bias, as discussed in the paper:

We recommend using either:
gaiaRVdelp54delsp43 : Increased parallax uncertainty by 0.043 mas in quadrature, corrected for parallax offset of 0.054 mas

gaiaRVdelpeqspdelsp43 : Increased parallax uncertainty by 0.043 mas in quadrature, corrected for a parallax offset = modified parallax error

For comparison and applications that need long-range in distance s, it may be interesting to use:
gaiaRVdelp48delsp00 : only increased parallaxes by an offset of 0.048 mas. Distance bias on average is stable out to 4kpc or more, but you have to be wary of random errors.


All distances were derived with the standard prior described in Schoenrich, McMillan and Eyer (2019); the parameter for the additional exponential function in the prior at very large s > 4 kpc is 0.06/kpc

If you need specific subsamples of stars, we advice to draw a new prior with the method of Schoenrich & Aumer (2017) and then re-derive the distances. We are also happy to help with that and provide you with the necessary code.

Please note that our definition (angular distance per time in the direction of l, b) of pml and pmb is frequently noted with a * in other derivations.

Please also note that why we used z_Sun = 0.02 kpc in the underlying analysis and distance prior, the value in the catalogue is given relative to the Sun (i.e. add the Solar offset that you favour).

Files

Files (5.1 GB)

Name Size Download all
md5:2c863336357ac9169a5e53524d9e8b3f
1.7 GB Download
md5:fb309b93a30ec38f2aec7f03b55c3739
1.7 GB Download
md5:55c53b3fa6bc8db6219747bfad2e24f9
1.7 GB Download