3. The proper motion of CoD-33°7795 B
CoD-33°7795 B was detected by L99 using HST NICMOS on 25 Apr 1998 in the F160W filter, located west 1 and north of CoD-33°7795 A, corresponding to a separation of and a position angle of . On 12 Jul 1998, Weintraub et al. (2000) detected the faint object, also using HST NICMOS, but with narrow band filters, located west and north of the bright star, corresponding to and . The precision in Weintraub et al. (2000) is higher than in L99, because the latter used the coronograph that makes it difficult to determine the centroid.
We present two new images of CoD-33°7795 B: A 1s exposure FORS2 I-band image taken during a technical night on 21 Feb 2000 with the high resolution collimator (/pixel) and a 2s exposure ISAAC acquisition image (/pixel) taken on 16 Apr 2000 through a narrow band filter centered on m (m). In both images, the central pixels of the bright star are saturated, which makes it difficult to determine the centroid; we fitted isophots in the unsaturated part of the PSF. The FWHM of the faint object on the FORS2 image is only , so that this image may well be the sharpest optical image ever taken from the ground (Fig. 3a).
In the FORS2 image, the companion candidate is located west and north of the bright star, corresponding to and , and in the ISAAC image, the companion candidate is located west and north of the bright star, corresponding to and . The errors include uncertainties in the north-south alignment.
In Fig. 3b, we plot the four positions of the companion candidate B with respect to star A with their error ellipses. If the error ellipses are disjunct by more than expected for orbital motion ( mas/yr, see below), object B could not be a co-moving companion. If the error ellipses do overlap, this does not prove object B to be a companion. Whether we can already show that the motion of CoD-33°7795 B relative to A is inconsistent with B being an unrelated field star, depends on the proper motion of star A. The proper motion was published by W99. In the Tycho catalog (Hog et al. 2000), we found and mas/yr.
In Fig. 3c, we plot star A first on 25 Apr 1998 at , then on 12 Jul 1998 south-west of it as given by its proper motion, and then on 21 Feb and 16 Apr 2000 even more south-west; the errors in the 2nd to 4th epoch locations are given by the error of the proper motion. In addition, we plot the offset of object B relative to star A with errors given by the errors of the measured offsets and the proper motion of star A. Object B is clearly co-moving with star A. If object B would be an unrelated field object, it should not be co-moving with A, but either be a non-moving background object or a foreground object with different motion (different parallactic motion would be negligible, even if unrelated, because the epoch difference between the HST and VLT images is close to an integer number of years). The error ellipses do not overlap. The proper motions of A and B are similar, namely by regarding their amount and by regarding their direction. Hence, we have in total a significance for the pair being a common proper motion pair.
© European Southern Observatory (ESO) 2000
Online publication: August 23, 2000