4.1. The spectra
The spatially integrated spectra of our objects are shown in Fig. 1. We extracted each spectrum from an aperture inside which the weakest lines were detected with higher S/N: MRC2025-218 (1.9 arc sec, centered at the spatial continuum centroid); MRC2104-242 (5.9 arc sec centered at the intersection between the two bright clumps); MRC1558-003 (2.7 arc sec centered at the spatial position of centroid of the brightest Ly component). The two spectra of SMM J02399-0136 correspond to a) the whole system (L1+L2) (the spectrum was extracted form a 5.4 arc sec aperture covering the brightest emission of the L1+L2 components) and b) he AGN component, L1. (2.2 arc sec aperture centered at the centroid of Ly emission in L1)
The spectra (except for SMM J02399-0136) are typical of HzRG. Ly is the strongest line and weaker CIV1550, HeII1640 and CIII]1909 are also detected. NV1240 is detected only in MRC2025-218 and SMM J02399-0136. This line is often not detected in HzRG (Röttgering et al. 1997). We compare in Fig. 2 the spectrum of L1 with an average spectrum of HzRG (Vernet et al. 1999) and the hyperluminous (also gravitational lensed) Seyfert 2 galaxy FSC10214+4724 (Goodrich et al. 1996). The differences are striking. L1 presents very weak HeII, strong NV and weak Ly compared to typical HzRG spectra. It is similar to FSC10214+4724 in the sense that Ly is weak and NV strong, but HeII is relatively much weaker in L1.
We present in Table 1 some parameters characterizing the main UV emission lines, resulting from 1-D Gaussian fits to the line profiles.
Table 1. CIV flux (in units of 10-16 erg s-1 cm-2). The fluxes of the strongest lines are given relative
4.2. The spatial distribution of the continuum and the emission lines
We described in VMFB99 the spatial properties of the Ly emitting gas derived from the 2-D spectra. Here we present 1-D spatial profiles for Ly, the continuum and the strongest UV lines. We created the emission line spatial profiles by adding those pixels (in the dispersion direction) where a given line is detected and subtracting the underlying continuum from a window of identical size (in Å) close to the line. The continuum spatial profile was created using a much larger window to increase the S/N ratio (typically 100-150 pixels in the dispersion direction, i.e. 65-90 Å). We present in Fig. 3 the spatial profiles for the strongest lines and the continuum.
Therefore, all objects present extended continuum over several arc sec (up to 15 arc sec in SMM J02399-0136). The continuum in MRC2025-218, MRC1558-003 and SMM J02399-0136 is dominated by a bright component which is rather compact. It is probably unresolved in SMM J02399-0136 and marginally resolved in MRC2025-218. All objects show extended emission lines which present rather different spatial profiles compared to the continuum (compare, for instance, the Ly and continuum profiles in SMM J02399-0136 and MRC2025-218). Both lines and continuum reveal the presence of several spatially distinct regions.
4.3. Absorption features in the spectrum of MRC2025-218
The 2-D spectrum of MRC2025-218 shows a clear absorption feature blueshifted with respect the CIV emission (see Fig. 4). In order to search for other absorption features, we have extracted a 1-D spectrum from the continuum emitting region (8 pixels or 2.2 arc sec aperture). We fitted the profiles of all possible absorption detections. We assumed Gaussian profiles (although it does not necessarily have to be the case). Some absorption features are detected. We show in Fig. 5 (bottom) the spectrum in the range 1180-1700 Å, with the expected position of some absorption features commonly found in nearby starburst galaxies. We present for comparison the spectrum of the B1 star forming knot in NGC1741 (Fig. 5 top) (Conti et al. 1996).
We rejected those features such that: a) the spectral profile was (taking errors into account) narrower than the instrumental profile (IP) (2.98 Å in the rest frame) and/or b) the absorbed flux was lower than the detection limit. This was the case of SiII1260.4, OV1371, SiIII1417, SV1502. In order to calculate the detection limit for an absorption feature at a given position, we created Gaussians with the expected FWHM (IP in all cases) and varied the amplitude (the profiles could be broader, but this just means that a larger flux would be needed for detection). The Gaussians were added to the continuum near the expected position. The upper limit was chosen by eye, as the flux of that Gaussian that we considered detectable.
We present in Fig. 6 the fits to those features that we accepted as real. Except for CIV (for which the original spectrum is shown), we present smoothed spectra to make the figures clearer. The fits were done to the original (non-smoothed) spectra. There are some sky emission residuals on the blue side of the CIV absorption feature, but they do not affect the fit (see Fig. 4). We present in Table 2 some parameters obtained from the fits: wavelength, line identification, EW (rest frame) and FWHM. The values measured in the radio galaxy 4C41.17 (3.80) (Dey et al. 1997) and the associated absorption system of the quasar 3C191 (1.95) (Bahcall et al. 1967) are also shown.
Table 2. Absorption lines measured in the spectrum of MRC2025-218. The EWs are given in the rest frame. Parameters of the emission line are given for those cases where a PCygni profile is detected. We present also the values measured for 4C41.17 (Dey et al. 1997) and 3C191 (Bahcall et al. 1967).
The fitting to SiIV1393,1402 is difficult due to the low S/N ratio, however, the presence of two P-Cygni profiles for the SiIV1393.8 and SiIV1402.8 lines is suggested by the data. The best fit is obtained by considering two absorption and two emission features. We constrain the fit so that the two absorption features on one hand, and the two emission features on the other, have the same FWHM and are separated by 9 Å (as is expected from the doublet components 1393,1402). The results of the fit are presented in Table 2. The FWHM of the emission line features are consistent with the value measured for the CIV and Ly emission. The absorption feature presents a much larger FWHM than measured for CIV absorption. The value is consistent with measurements in galactic outflows.
The absorption features are narrower than the emission lines (also narrower than in 4C41.17) except for the SiIV1393,1402 lines, that have FWHMabs 1400500. The absorption features in the P Cygni profiles are blueshifted by 1100200 km s-1 (CIV) and 1200200 (SiIV1393.8 and SiIV1402.8) with respect to the emission.
We confirm the detection in absorption of CIV1550, CII1334.5, SiIV1393.8 +1402.8 and, maybe, OI1302.2 +SiII1304.4 (the absorbed flux is slightly higher than the detection limit). There is no clear evidence in our data for Ly absorption, although the asymmetry of the profile (steeper on the blue side) is probably due to absorption. The presence of P Cygni profiles is confirmed in the case of CIV, SiIV1393.8 and SiIV1402.8.
© European Southern Observatory (ESO) 1999
Online publication: November 2, 1999