Erratum for: A&A 628, A14 (2019) https://doi.org/10.1051/0004-6361/201935064

The polarization angles that are reported in Lankhaar & Vlemmings (2019) have to be revisited. It has been pointed out to us that a bug in our code lead to the wrong sign of the Stokes parameters

$$\begin{array}{cc}& Q(\omega )\to -Q(\omega )\hfill \\ \hfill & U(\omega )\to -U(\omega ).\hfill \end{array}$$

This error has no influence on the total maser brightness, the total polarization fraction, or the circular polarization, but it leads to an erroneous polarization angle. The polarization angle that we report in Lankhaar & Vlemmings (2019), χ, is related to the correct polarization angle, χ′, by

$$\begin{array}{c}\hfill \chi \to {\chi }^{\prime }=\{\begin{array}{cc}\chi -\pi /2\hfill & \mathrm{for}\chi \ge 0\hfill \\ \chi +\pi /2\hfill & \mathrm{for}\chi <0.\hfill \end{array}\end{array}$$(1)

The error has lead to erroneous statements in our discussion in the paragraph beginning with “90^o flip of the polarization angle.” We state that when gΩ ≳ 100R and propagation angles θ <  θ_m, the polarization is oriented perpendicular to the projected magnetic field direction and for θ >  θ_m the polarization is oriented parallel to the projected magnetic field. Using Eq. (1), this statement should be the other way around: When gΩ ≳ 100R and propagation angles θ >  θ_m, the polarization is oriented perpendicular to the projected magnetic field direction and for θ <  θ_m the polarization is oriented parallel to the projected magnetic field. The polarization flip we discuss is still present. We also state in this paragraph that in the case R ≫ gΩ, the polarization is oriented parallel to the projected magnetic field direction. This should be corrected according to Eq. (1): For R ≫ gΩ, the polarization is oriented perpendicular to the projected magnetic field direction.

In Figures (1c), (2b), (3), (4b,d), (7b), (8b), and (9), the polarization angles that we plotted have to be transformed according to Eq. (1). The same goes for the appended figures (A1b,e), (A2b,e,h), (A3b,e,h), (A4b,e), (A5b,e,h), (A6b,e,h), (A7b,e), (A8b,e,h), (A9b,e,h), (A10b,e), (A11b,e,h), (A12b,e), (A13b,e), (A14b,e), (A15b,e), (A16b,e), (A17b,e), (A18b,e), (A19b,e,h), (A20b,e,h), (A21b,e,h), and (A22b,e,h).

Acknowledgments

We thank Gabriele Surcis for pointing out the wrongly predicted polarization angles to us.

References

© ESO 2020