We analyse the sizes, colour gradients, and resolved stellar mass distributions for 36 massive and passive galaxies in the cluster XMMUJ2235-2557 at z=1.39 using optical and near-infrared Hubble Space Telescope imaging. We derive light-weighted Sersic fits in five HST bands ($i_{775},z_{850},Y_{105},J_{125},H_{160}$), and find that the size decreases by ~20% going from $i_{775}$ to $H_{160}$ band, consistent with recent studies. We then generate spatially resolved stellar mass maps using an empirical relationship between $M_{*}/L_{H_{160}}$ and $(z_{850}-H_{160})$ and use these to derive mass-weighted Sersic fits: the mass-weighted sizes are ~41% smaller than their rest-frame $r$-band counterparts compared with an average of ~12% at z~0. We attribute this evolution to the evolution in the $M_{*}/L_{H_{160}}$ and colour gradient. Indeed, as expected, the ratio of mass-weighted to light-weighted size is correlated with the $M_{*}/L$ gradient, but is also mildly correlated with the mass surface density and mass-weighted size. The colour gradients $( abla_{z-H})$ are mostly negative, with a median value of $sim0.45$ mag dex$^{-1}$, twice the local value. The evolution is caused by an evolution in age gradients along the semi-major axis ($a$), with $ abla_{age} = d log(age) / d log(a)$ $sim-0.33$, while the survival of weaker colour gradients in old, local galaxies implies that metallicity gradients are also required, with $ abla_{Z} = d log(Z) / d log(a)$ $sim-0.2$. This is consistent with recent observational evidence for the inside-out growth of passive galaxies at high redshift, and favours a gradual mass growth mechanism, such as minor mergers.
Download