All Figures

	Figure 1: The (u,v)-coverage obtained on 2050+364 with the EVN telescopes Effelsberg, Jodrell Bank, Onsala, and WSRT. Crossing points provide important self-calibration constraints.
Open with DEXTER


In the text

	Figure 2: The dirty beam obtained with the (u,v)-coverage shown in Fig. 1; contour levels are drawn at -0.2, -0.1, -0.05, 0.05, 0.1, 0.2, 0.4, and 0.8. This beam is benign for deconvolution of the continuum structures in Fig. 3.
Open with DEXTER


In the text

Figure 3: Continuum VLBI image of 2050+364 at 1049 MHz. The restoring beam is 50  38 mas in position angle -27. The peak flux density is 2.42 Jy/beam; contours start at 0.015 Jy/beam (3 times the rms fluctuation level away from the continuum structure), and increase in factors of 2. The absolute flux density scale is uncertain by 20%, but this does not affect relative flux densities; see Sect. 2.2. The three circular Gaussian components used for modeling (Table 1) are overplotted.

Open with DEXTER


In the text

	Figure 4: The H I absorption spectra, expressed in opacity, for the W, E, and N radio components of 2050+364. Zero velocity is chosen to be the [O III] redshift, z=0.3547 (see Sect. 5.2).
Open with DEXTER


In the text

	Figure 5: Enlargement of part of Fig. 4, to allow detailed comparison of the H I opacity profiles towards W and E.
Open with DEXTER


In the text

	Figure 6: Fitted Gaussian line profiles overplotted onto the H I absorption spectra of the three regions. The parameters are listed in Table 2.
Open with DEXTER


In the text

	Figure 7: Continuum component flux densities at multiple frequencies, based on data at 0.33, 0.61, 1.67, 2.3, and 8.4 GHz from LF, at 2.3, 8.4, and 15.4 GHz from K03, and at 1.049 GHz from our own observations. See Sect. 4 for details.
Open with DEXTER


In the text