Correction to: The PHANGS–MUSE nebular catalogue

This is a correction to Groves et al. (2023).

In table 1 of the original paper, we note that the footnotes (f) and (g) are switched.

In section 4.4, equation (4) had a mistake in the sign, however it did not affect the calculations done for the analysis of the paper. The coefficient was written as |$-1.684$|⁠, however it should be |$+1.684$|⁠. The corrected equation, that is based on the prescription of Diaz et al. (1991) to determine U from a diagnostic line ratio, is

There was also a mistake in calculating the values of log(U) that are used in figs 7, 9, 11 and appendix fig. C3. In the literature, the ionization parameter is described by one of two parameters, q or U, which can be related as |$U\!$| = q/c, where c is the speed of light. When converting from an initial calculation of the parameter q to U for the paper, we divided by the speed of light (c) using mks units instead of cgs units. Due to this mistake, all values of log(U) are too large by a factor of 100. With the correctly calculated values of log(U), we see that our H ii regions cover a range of ionization parameters from −5.0 |$<$| log(U) |$<$| −2.5. All conclusions drawn in the paper are related to relative comparisons of log(U) within the sample, so none of the conclusions in the paper are affected by this mistake. The values in the machine readable table containing the nebular catalogue and associated with this paper are correct.

The corrected version of fig. 7 can be seen in Fig. 1.

Figure 1.

Metallicity (12 + log(O/H)) as a function of ionization parameter (U) for H ii regions within each galaxy. Galaxies are sorted by stellar mass from low (top left) to high (bottom right). Linear fits are performed in each galaxy, demonstrating the prevalence of positive correlations across the sample. Apart from the well-established mass–metallicity relation, no obvious trends are seen as a function of total galaxy stellar mass (red colour scale, bottom right).

Open in new tabDownload slide

The corrected version of fig. 9 can be seen in Fig. 2.

Figure 2.

Radial gradients for four representative galaxies in key ISM diagnostics (from top to bottom): H |$\alpha$| luminosity (L(H |$\alpha$|⁠)), extinction derived from the Balmer decrement (A|$_V$|⁠), metallicity (12 + log(O/H)), and ionization parameter (U). For each galaxy we show the radial trends normalized to r|$_{\rm eff}$|⁠. These four galaxies represent low stellar mass (IC 5332) and high stellar mass (NGC 1365) galaxies, and systems with a regular spiral pattern and no bar (NGC 0628) or strong bar and widely separated arms (NGC 7496). Most trends are flat or mildly negative, except for the metallicity gradients which show the strongest negative trends.

Open in new tabDownload slide

The corrected version of fig. 11 can be seen in Fig. 3.

Figure 3.

Comparison of median extinction (A|$_V\!$|⁠; top) and median ionization parameter (U; bottom) as a function of global galaxy properties (stellar mass, star formation rate, gas fraction). In the top left of each plot we show the Pearson correlation coefficient of those quantities, r. A|$_V\!$| shows modest correlations with stellar mass and SFR, reflecting an increased gas content (and hence dust content) in these systems. |$U\!$| shows no correlations, reflecting that it is set primarily by local ISM conditions.

Open in new tabDownload slide

The corrected version of fig. C1 can be seen in Fig. 4

Figure 4.

Ionization parameter (U) radial gradients.

Open in new tabDownload slide

The corrected version of table C1 can be seen in Table 1.

In table 7 of the original paper, we provided conversion formulae between different metallicity conventions, but they do not have enough significant digits to be useful. They are also not easy to interpret, because the intercepts were calculated at 12 + log(O/H) = 0 whereas most data points span [8, 9]. To fix this, we now provide in Table 2 the intercepts at a value of 12 + log(O/H) = 8.6 (as this is representative of the median of the data range) and increase the number of significant digits shown. The fits shown in the original fig. 5 remain unchanged.

REFERENCES