3. OH and HCO+ profiles, then and now
3.1. Stability of HCO+ absorption over 3-5 year intervals
New and old profiles of HCO+ absorption in the 78 kHz channel spectra are shown and differenced in Figs. 1 and 2. Profiles of the line optical depth and optical depth differences are shown for three sources in Fig. 3.
In several cases, small imperfections in the spectral baseline are manifested across the difference specta. For components as broad as some of those seen here this could be easily misconstrued. But the red wings of the -17 km s-1 and -4 km s-1 components toward B0355+508, the entire stronger line toward B0528+134, and the central region of the deep line toward B2200+420 have statistically significant differences. The -4 km s-1 component toward B0355+508 noticeably changes its shape. All the variation toward this source represent changes in the profile integrals. Toward B0528+134 and B2200+420 the variations have the characteristic, partially integral-conserving up-down shape of a simple shift in velocity (though we cannot be certain that this is in fact what has occurred).
HCO+ profile integrals are presented in Table 3, where the column labelled `Old' gives the epoch at which the older data were finalized. Differences between our old and new values are small and usually of little significance. The total profile integrals summed over all components are stable at the level of 1% in the best cases. For B0355+508 there is a pattern whereby the old integrals for individual components are greater toward the blue end and smaller to the red; this pattern is not repeated in OH, suggesting that it is due to some unrecognized systematic effect. The difference in profile integrals toward B0415+379 is significant in terms of the channel to channel rms error but is 1.7% overall. Evidence for variation in the OH toward this source is equivocal.
3.2. OH, then and now
OH profile integrals for are given in Table 2 along with results of (Moore and Marscher 1995) and (Marscher et al. 1993) (labelled `BU') courtesy of A. Marscher who kindly forwarded his profiles. As noted in (Liszt and Lucas 1996), these BU data were actually used to normalize our old profile toward B0355+508; the profile integrals at v km s-1 were set equal, so that only the distribution of absorption among the various components may be compared. The BU data agree very well with ours for the individual components, suggesting that some of the time-variability between our old and new data is real. In particular, the -10.4 km s-1 line changes by . But this component of the HCO+ shows only a slight, opposite change.
The total profile integrals differ significantly between our old and new data in a few cases, most notably B2200+420 and B0415+379. Yet in each case, the BU data are closer to our newer results and the observed variations are not repeated in the HCO+ profiles.
3.3. Comparison of HCO+ and OH profiles
This comparison is illustrated in Figs. 4 and 5 where we have superposed the HCO+ and OH optical depth profiles (without any adjustment of the velocity axes). The OH has been normalized to the same total area, except for B0212+735 and B1730-130 where only the area under the stronger line was set equal. The velocity resolution for the newer HCO+ data is 0.23 km s-1 with 0.131 km s-1 channel separation but two of the HCO+ profiles in the comparison are older and have lower resolution; for B0212+735 we have only the 0.47 km s-1 resolution data from our earlier epoch and toward B0415+379 the comparison is made with H13CO+ (since the main HCO+ line is so strongly saturated) for which the velocity resolution is 0.49 km s-1. The OH spectra were convolved down to the appropriate resolution but kept at their original channel separation of 0.137 km s-1.
In general, the profiles of these two species are very similar as befits their tight correlation of abundance. The extent of both species is sensibly identical but the -21 km s-1 line of OH toward B0355+508 is very weak in HCO+ and the +7 km s-1 red wing of the HCO+ line toward B1730-130 is absent in OH. There is a somewhat narrower line core in OH toward BL Lac and perhaps also in the stronger line toward B0212+735 (where the spectral resolution of the HCO+ is low). The blue wing of the profile toward B0415+379 is relatively weak in OH when compared to either HCO+ isotope, suggesting a partial kinematic decoupling of these (usually) chemically tightly-linked species. If there are other means of forming CO in a regime where H chemistry is well-developed, there are routes to HCO+ formation which do not involve OH. But the OH/HCO+ ratios are entirely typical toward 3C111. In (Lucas and Liszt 1998) we noted that the carbon isotope ratios in the blue component are unusual in some species.
Single-component Gaussian fits to the lines do not show a clear pattern of differences in the line widths; for the strong lines toward B0212 and B1730, and for B1954, HCO+ is 4%-8% broader (FWHM linewidths of 1.18, 0.77, and 0.77 km s-1 for HCO+ vs 1.13, 0.71 and 0.72 km s-1). Toward B0355, HCO+ in the three stronger lines is narrower (0.61, 0.51, and 1.07 km s-1 vs 1.02, 0.58, and 1.08 km s-1). Toward B2200, a one component fit gives a much broader HCO+ line, 1.47 km s-1 vs 1.24 km s-1 but the HCO+ line is really compound, composed of two slightly-separated components of nearly equal strength which simply cannot be well-decomposed by Gaussian fitting. The weaker line toward B0212+735 is narrower in HCO+, 0.75 vs 0.94 km s-1.
Sadly, we note briefly that the HCO+ and OH lines are too narrow and too similar to be explained by profiles such as those shown for C-shock models by (Flower and Pineau Des Forêts 1998), which otherwise do a good job in reproducing the correlation of abundances. The double nature of the OH profiles in these models, arising from contributions from both the host (quiescent) and shocked gas, is not seen. In most cases, it is the HCO+ which has the wider lines in absorption. However, it should be noted that the optical depth of OH is very sensitive to its excitation and this is not understood at the moment (Liszt and Lucas 1996): OH excitation temperatures should generally be much higher than they are (typically, only 1 K above the 2.7 K background) with a consequent reduction in the line optical depths.
© European Southern Observatory (ESO) 2000
Online publication: March 17, 2000