Robin Catchpole, Alec Boksenberg and the FOC team
Royal Greenwich Observatory, Madingley Road, Cambridge CB3 0EZ, UK.
NGC 1068 (M77) is the closest and brightest example of a Seyfert 2 galaxy. Many active galactic nuclei (AGN), including NGC 1068, have apparently conical or biconical high-excitation emission line structures extending from the position of the nucleus (e.g., Pogge 1988, Tadhunter & Tsvetanov 1989, Unger et al. 1992), which are taken as evidence of anisotropy and orientation effects evoked to explain the perceived differences between the various observationally classified AGN types. In this unifying hypothesis radiation is presumably collimated by an optically thick torus surrounding the nucleus. Whether a Seyfert galaxy of type 1 or 2 is seen depends on the orientation of the torus to the line of sight.
In initial HST Planetary Camera (PC) observations of the nuclear region of NGC 1068 an [O III] 5007 image was obtained by Evans et al. (1991) and an optical continuum image by Lynds et al. (1991). Macchetto et al. (1994) used the COSTAR-augmented Faint Object Camera (FOC) to observe the nuclear region at a number of wavelengths. Capetti et al. (1995a,1995b) sought to locate the position of the nucleus using pre- and post-COSTAR polarisation data. Boksenberg et al. (1996) in their discussion of the HST images note the presence of a system of rays or shadow lines (crepuscular rays) suggesting the presence of a bright nuclear source but at a location different from those found by Capetti et al. (1995a,1995b).
The continuum radio map at 4.9 GHz obtained by Wilson & Ulvestad (1983) shows a double-lobed radio jet centered on the nucleus with a projected length of 6 to the north-east and about 10 to the south-west. Muxlow et al. (1996) have identified a faint source in their 5 GHz 60 mas resolution map which they believe to be the nucleus of NGC 1068 on the basis of its spectral slope.
Both radio and optical observations show structures elongated along PA . One of the outstanding questions regarding NGC 1068, which we address here, is the matching of the HST and radio images and hence the location of the nucleus in the HST images.
The 13 and
plate material used to determine positions published by Clements (1981),
has been re-measured using a PDS machine and reduced by J. Pilkington and
R. Argyle of the Royal Greenwich Observatory. Their reduction used reference
star positions obtained with the Carlsberg Automatic Meridian Circle (CAMC)
which are tied to the FK5 reference frame.
They obtain a position (equinox 2000.0) for the optical centre of NGC 1068:
2 42 40 0 -0^o 0 47 0
The reference frame of the Muxlow et al. (1996) 5 GHz map of NGC 1068 is
tied to the IERS reference frame with an uncertainty of order 0
The offset between the ICRF system (essentially identical to the IERS) and
the FK5 as defined by the CAMC,
has been determined by Argyle et al. (1996) and is given by:
In Figure 1 we include these corrections and show the astrographic position in relation to the 5 GHz map. The error circle includes the quadratic sum of the transformation error between the two systems and the internal errors in the measured positions of NGC 1068.
The remaining and greatest uncertainty lies in relating the `nucleus' of NGC 1068, measured on the astrographic images, to the HST images. The astrographic position was found by first using the PDS machine to raster scan a 0 square aperture over the photographic images. The centre of each image was then found by fitting a circularly symmetric Gaussian profile to the measured distribution in photographic transmission.
Figure: The Muxlow et al. (1996) 5 GHz map of NGC 1068 showing the position of the centre measured here on astrographic plates. The error circle includes both the error of measurement and the uncertainty in the transformation from the FK5 to the IERS reference frame. The equinox of the grid is 2000.0. Also shown is the brightest part of the FOC F550M (left) and F501N (right) post-COSTAR images, which are related to the radio reference frame as discussed in the text. The position of the origin of the crepuscular rays (Boksenberg et al. 1996) and the Braatz et al. (1993) 12.4 source are located relative to the OCP. These error bars do not include the 0 uncertainty in relating the HST frame to the radio reference frame.
The observed FWHM for a faint star image in photographic transmission, is about 4 . We have simulated this by blurring the post-COSTAR F501N and F550M images using a Gaussian profile with FWHM of 2 and 4. The centres of the resulting images were then found by fitting Gaussian profiles within an aperture of 2. The final positions, corresponding to the two blurrings, were found by taking a weighted mean of the F501N and F550M positions, where the weighting simulates the wavelength sensitivity of the Kodak IIaO emulsion. The resulting positions lie along PA = 48^o, 0 and 0 from the optical continuum peak (OCP) for the 2 and 4 blurrings, respectively. The dependence of the peak in the flux on the degree of blurring indicates that we can not match the HST image to the astrographic position with a precision better than about 0. We fit the HST to the 5 GHz map using the mean position of the 2 and 4 blurrings. This is shown in Fig. 1, where we also show the Braatz et al. (1993) 12.4 source, which they locate with respect to the OCP. It is interesting to note that not only does the error box of the 12.4 source include the origin of the crepuscular rays but the fit to the radio map places the radio nucleus in the same vicinity. This strongly suggests that all three sources are coincident and that the nucleus of NGC 1068 is indeed hidden from direct view in a region of strong obscuration.
Our adopted position for the nucleus is significantly different from that of Capetti et al. (1995a) who placed it along PA = 170^o, 0 from the OCP, and Capetti et al. (1995b) who placed it along PA = 168^o, 0 from the OCP, on the basis of their respective analyses of pre- and post-COSTAR polarisation data.
Figure: The post-COSTAR FOC F501N (left) and F253M (right) images are compared with the 5 GHz map when the OCP is brought into coincidence with the brightest radio source. The HST image area is identical to Fig. 1.
It is clear from Fig. 1 that the uncertainties in the fit are much greater than the sizes of the optical and radio structures. The uncertainties are large enough to allow either the OCP to coincide with the brightest radio source or the radio nucleus to coincide with the origin of the crepuscular rays. The first option, shown in Fig. 2, places the radio nucleus closer to the 12.4 source. It implies that the radio jet changes direction at the OCP and then passes through a region bounded on either side by line emission. In this case the axis of the UV cone is expected to lie along the initial direction of the radio jet (PA 13.1) implying that it is a coincidence that the subsequent alignment of the radio jet and the large scale UV cone should be almost identical. The possibility that the UV cone originates at the OCP and not the nucleus would be in disagreement with the unified model for AGN.
If the radio nucleus coincides with the origin of the crepuscular rays then there is no detailed correspondence between the radio and optical features. Apart from the evident nucleus, this is the situation seen in NGC 4151 (Boksenberg et al. 1995). If so, then the axis of the UV cone, originating from the nucleus, could lie either along PA 13.1 or along PA . In the second case the axis would pass between the bright radio source and the OCP.
We are indebted to R. W. Argyle and J. D. H. Pilkington for their re-measurement of the Clements (1981) data. The CAMC is operated jointly by Copenhagen University Observatory, the Royal Greenwich Observatory and the Real Instituto y Observatatorio de la Armada en San Fernando.
Argyle, R. W., Einicke, O. E., et al. 1996, A&A, submitted
Boksenberg, A. B., Catchpole, R. M., et al. 1996, to be published
Boksenberg, A. B., Catchpole, R. M., Macchetto, F., et al. 1995, ApJ, 440, 151
Braatz, J. A., Wilson, A. S., Gezari, D. Y., Varosi, F., & Beichman, C. A. 1993, ApJ, 409, L5
Capetti, A., Axon, D. J., Macchetto, F., Sparks, W. B., & Boksenberg, A. 1995a, ApJ, 446, 155
Capetti, A., Macchetto, F., Axon, D. J., Sparks, W. B., & Boksenberg, A. 1995b, ApJ, 452, L87
Clements, E. D. 1981, MNRAS, 197, 829
Evans, I. M. et al. 1991, ApJ, 369, L27
Lynds, R. et al. 1991, ApJ, 369, L31
Macchetto, F., Capetti, A., Sparks, W. B., Axon, D. J., & Boksenberg, A. 1994, ApJ, 435, L15
Muxlow, T. W. B. 1995, private communication
Muxlow, T. W. B., Pedlar, A., Holloway, A.J., Gallimore, J. F., & Antonucci, R. R. J. 1996 MNRAS, in press
Pogge, R. W. 1988, ApJ, 328, 519
Tadhunter, C. & Tsvetanov, Z. 1989, Nature, 341, 422
Unger, S. W., Lewis, J. R., Pedlar, A., & Axon, D. J. 1992, MNRAS, 258, 371
Wilson, A. S. & Ulvestad, J. S. 1983, ApJ, 275, 8