Erratum: The large-scale environment from cosmological simulations I. The baryonic cosmic web

This is an erratum to our paper ‘The large-scale environment from cosmological simulations I. The baryonic cosmic web’, which was published in MNRAS 473, 68 (2018). Here, we correct an error in both Figs 4 and 5 of the mass fractions, which have inverted x-axes in the original publication. We provide below the corrected plots, but also note that several sentences need to be revised to properly reflect them, which are shown in bold font:

1. The last paragraph on page 70 should read as “The variations are better seen in the lower panel of Fig. 4, which shows the quantitative difference between the respective fractions runs with respect to the DM-only run. For the VWEB results, the AGN run tends to have both a slightly larger (∼2 per cent) volume and mass fraction in the highest density region (i.e. knots), while the CSF run gives ∼1 per cent lower volume and mass fractions than the DM-only run. Without AGN feedback, the CSF run tends to have more concentrated knots with a relatively weaker velocity field, which tends to occupy less spatial volume and mass; while the strong AGN feedback not only stops star formation but also pushes matter into outer regions, which results in a large volume with a higher velocity field (see discussion in Ragone-Figueroa et al. (2013); Cui et al. (2014, 2016). In filament, sheet and void regions, both mass and volume fractions show almost no change between these three runs. As the PWEB code directly uses the second derivatives of the potential, which are directly connected to the density via Poisson’s equation, to classify these structures, there is even less difference (≤1 per cent) between the two hydrodynamical runs and the DM simulation for all structures.”
2. The first item of the itemize on page 73 should read as “(1) In knots, only ∼90 per cent of the cells in the AGN run are consistent with the results from all matter (∼93 for the CSF run). And the mass fraction for both the CSF and AGN runs are ∼4 per cent lower. The knot volume fraction in the CSF run is about ∼4 per cent higher, while the AGN volume fraction shows almost no change.”
3. The last sentence of item 3 on page 73 should read as “(3) In void regions, the volume and consistency fractions are basically within 1 per cent for both runs. However, the mass fraction tends to be ∼4 per cent higher for both CSF and AGN runs.”
4. The first paragraph on page 74 should reads as “These differences agree well with the expectations and the influence of baryonic processes: In knots, baryon processes are more violent. Thus, the velocity field derived from gas can deviate from the total velocity field, which results in a relatively lower mass fraction and about 10 per cent inconsistency fraction for both the CSF and AGN runs. In void environment, gas tends to have higher mass fractions although about 99 per cent of these void cells are in agreement with the all matter results. This indicates a slightly higher gas density contrast in the void regions compared to the DM one.”

None of our final conclusions were affected by this error.