Accurate predictions for the scale-dependent galaxy bias from primordial non-Gaussianity

The large-scale clustering of galaxies can serve as a probe of primordial non-Gaussianity in the Universe competitive with the anisotropies of the cosmic microwave background. Here, we present improved theoretical predictions which include an important, previously overlooked correction to the bias. We demonstrate that the new predictions are able to reproduce the results of N-body simulations, explaining the significant departures seen from previous theoretical results. These refined predictions open the way to accurate constraints on primordial physics with large-scale structure surveys.

Figure 1

Ratio of the large-scale NG halo bias measured in simulations with non-Gaussianity of the scale-dependent local type ($n_f=\pm 0.6$) [8], to that predicted by the high-peak approximation. Filled and open symbols show the simulation results, whereas the solid and dashed curves indicate the new theoretical expectation [Eq. (10)] at $z=0$ and 1, respectively.

Figure 2

Same as Fig. 1 but for the orthogonal bispectrum shape [9]. Different symbols correspond to various halo mass bins as indicated in the figure; colors indicate the redshifts $z=0$ (magenta), 0.67 (blue), 1 (green) and 1.5 (red). The solid curve shows the new theoretical prediction assuming a halo mass $M=5\times {10}^{13}{M}_{\odot }/h$.

Figure 3

Same as Fig. 1 but for the local cubic model. Triangle and square symbols indicate results for $g_{\mathrm{NL}}=-{10}^6$ and ${10}^6$, respectively. Different colors show different redshifts spanning the range $0<z<2$.