Halo assembly bias and the tidal anisotropy of the local halo environment

We study the role of the local tidal environment in determining the assembly bias of dark matter haloes. Previous results suggest that the anisotropy of a halo's environment (i.e. whether it lies in a filament or in a more isotropic region) can play a significant role in determining the eventual mass and age of the halo. We statistically isolate this effect, using correlations between the large-scale and small-scale environments of simulated haloes at z = 0 with masses between 10(11.6) less than or similar to (m/h(-1) M-circle dot) less than or similar to 10(14.9). We probe the large-scale environment, using a novel halo-by-halo estimator of linear bias. For the small-scale environment, we identify a variable alpha(R) that captures the tidal anisotropy in a region of radius R = 4R(200b) around the halo and correlates strongly with halo bias at fixed mass. Segregating haloes by aR reveals two distinct populations. Haloes in highly isotropic local environments (alpha(R) less than or similar to 0.2) behave as expected from the simplest, spherically averaged analytical models of structure formation, showing a negative correlation between their concentration and large-scale bias at all masses. In contrast, haloes in anisotropic, filament-like environments (alpha(R) less than or similar to 0.5) tend to show a positive correlation between bias and concentration at any mass. Our multiscale analysis cleanly demonstrates how the overall assembly bias trend across halo mass emerges as an average over these different halo populations, and provides valuable insights towards building analytical models that correctly incorporate assembly bias. We also discuss potential implications for the nature and detectability of galaxy assembly bias.