6. Constraints on the quasar CO source
In addition to the size of the quasar CO source, 150-460 pc FWHM, as derived above (see Sect. 5.2), we can obtain some information on its substructure using the velocity gradient observed in the CO(7-6) line (Fig. 1b). Indeed, the map (Fig. 1c) corresponding to the blue side of the line (-225,-25 km/s) shows that spot C is stronger and slightly displaced inwards with respect to its counterpart in the map corresponding to the (+25,+225 km/s), shown in (Fig. 1d). A hint of this effect can be found in Yun et al (1997) although a close comparison of the IRAM results with the OVRO one is hampered by the fact that in the OVRO study the spatial resolution is twice lower, the line profile is not shown, the velocity interval considered - 145 km/s - is narrower than ours - 200 km/s - and not positioned precisely with respect to the line center.
Therefore, we shall concentrate now on the interpretation of the IRAM data set only. Using the two lens models obtained in Sect. 5.2, we optimized the structure of the CO source to reproduce separately the red and the blue images (Fig. 7). We call attention to the fact that a precise registration (better than ) of the CO map and the HST data is required to derive a detailed structure of the CO source. The current data does not allow such an accurate registration, so we fine tune it so that the lens morphology between radio and optical match is best. We tested the registration and found that a offsets does not change the results presented below. A larger offset make the radio and optical data inconsistent on a lens modelling point of view.
For model 1 [resp. model 2], we find a difference of ( pc) [resp. ( pc)] between the center of regions emitting the blue and red parts of the CO line (c.f. Fig. 7). We have tested this procedure against uncertainties in the CO/HST images registration. A change by (half a CO map pixel) has a very minor impact on the positions of the source regions emitting the blue and red parts of the line. The quasar point-like visible source appears to be almost exactly centered between the blue- and red-emitting regions. This is reminiscent of a disk- or ring-like structure orbiting the quasar at a radius of 100 pc [resp. 75 pc] and with a Keplerian velocity of 100 km/s (assuming a 90 deg inclination with respect to the plane of the sky), the resulting central mass would be 109 M [resp. 7.5 108 M ]. Elaborating a more sophisticated (realistic) model of the CO source and its link with the BAL feature also observed in this quasar is beyond the current analysis but should be performed in the future. Regarding the spot fluxes in CO(7-6) as derived from the maps in Figs. 1a, 1c, 1d and Table 1, a comparison with the fluxes given previously in Alloin et al (1997) is not straightforward because the velocity interval over which the integration has been performed is different in the two studies (-225,+225 km/s vs. -325,325 km/s). The amplification factor being extremely sensitive to the position of the emitting region with respect to the caustic, any velocity-positional changes within the source can result in a different configuration in the image plane. Such gradients are likely to occur in the molecular torus of the standard AGN model. The differential amplification resulting from this gradient is probably at the origin of the asymmetry observed in the line profile (Fig. 2).
© European Southern Observatory (ESO) 1998
Online publication: December 16, 1997