Extended Data Fig. 1: Slices of phase space used to isolate Gaia–Enceladus stars.

a, Energy E versus angular momentum L_z for stars in the 6D Gaia dataset that satisfy the quality criteria described in the text, with ϖ > 0.2 mas (5 kpc from the Sun) and |V − V_LSR| > 210 km s⁻¹. The dashed lines indicate the criteria used to select Gaia–Enceladus stars, namely, −1,500 kpc km s⁻¹ < L_z < 150 kpc km s⁻¹ and E > −1.8 × 10⁵ km² s⁻². These criteria follow roughly the structure’s shape (for comparison, see Extended Data Fig. 3b) but are slightly conservative for the upper limit of L_z to prevent too much contamination by the thick disk. However, small shifts—such as those obtained by considering an upper limit of 250 kpc km s⁻¹ or a lower limit of −750 kpc km s⁻¹ for L_z, or E > −2 × 10⁵ km² s⁻²—do not result in drastic changes to the results presented in the paper. The colour scale indicates the logarithm of the counts in the bins, with red corresponding to the highest number of counts, yellow and blue to 1/6th and 1/30th of this value, respectively, and purple to empty bins. b, Angular momentum L_z versus the Galactocentric distance R for all stars in the 6D Gaia with ϖ > 0.2 mas. The black points are the halo-star sample shown in a. c, Same as b, but for star particles in the merger simulation¹⁸ shown in Fig. 1b. Blue points correspond to stars from the satellite and grey points to the host disk, and the positions and velocities have been scaled as described in the text. In this figure, E, L_z and R have been scaled by the energy (E_sun = −1.63 × 10⁵ km² s⁻² in the Galactic potential used), angular momentum (L_z,sun = 1,902.4 kpc km s⁻¹) and Galactocentric distance (R_sun = 8.2 kpc) of the Sun, respectively.

Source data