Astron. Astrophys. 329, 827-839 (1998)

4. Additional constraints from the absolute registration of images in various wavebands

In order to derive precisely the shear induced by the gravitational lens on an extended source in the quasar, it is imperative to register with a high accuracy the Cloverleaf image in a waveband corresponding to point-like images of the quasar (i.e. R-band corresponding to rest wavelength 1967 Å) and in a waveband corresponding to an extended source in the quasar (i.e. the CO(7-6) molecular emission). The high precision required, better than , cannot be achieved from the HST data because of their limited field of view. Hence, we have searched for wide-field exposures of the Cloverleaf field obtained under good seeing conditions at CFHT.

4.1. The FOCAM CFHT data set

We have retrieved from the CFHT data archive (CADC database in Victoria) three images obtained on 1992 February 27 by Angonin, Vanderriest and Chatzichristou, with FOCAM. The CCD in use was Lick II (1500 pixels of ), covering a field. Two images were obtained through the R band and one through the I band, under a seeing condition. For every image, the integration time was 310 seconds. To derive the astrometry, we used the two R band direct images, unprocessed since we are only interested in the astrometric information.

4.2. Absolute astrometry

The absolute astrometry was performed using 6 reference stars across the field, from the Cambridge APM database (much deeper than the Guide Star catalogue). This allows absolute positioning of the brightest spot A within an accuracy of (rms). Then, we have used the relative astrometry of spots B, C and D, with respect to A, from the HST image (relative accuracy ). The absolute astrometry of the Cloverleaf is provided in Table 3 (J2000.0).

4.3. Registration of the optical and millimeter images

The IRAM CO(7-6) image is obtained within an absolute astrometric accuracy of (rms), quite comparable to that achieved from the combined CFHT/FOCAM and HST/WFPC2 results. Therefore, it is possible to register both the optical and the millimeter data in an absolute manner (Fig. 1). This is the starting point of our new modelling of the Cloverleaf.

In summary, we consider for the modelling the positions as well as R and I intensity ratios (the least affected by dust absorption) of the four spots given in Table 4, since we know that they are not affected by any spatial resolution effect. Conversely, the relative positions of the CO spots with respect to the visible spots are subject to some spatial resolution effect because the interferometer natural beam size is of the order of the spot size. Therefore, we shall not use the CO spot positions to build the lens model, but rather check whether the CO spots predicted by the lens model and convolved by the interferometer beam reproduce the observed offsets between the CO and the visible spots (see Sect. 5 and 6)

© European Southern Observatory (ESO) 1998

Online publication: December 16, 1997
helpdesk.link@springer.de