back to the 2003 results page

Deflotte on Jan. 23 : probable ~0".2 South shift

Jean Lecacheux (2003 feb. 18)

The path of 1295 DEFLOTTE on last January 23th was first predicted by E.Goffin, on basis of Tycho-2 star position, from Tarragona (e.Spain) to the north border of Portugal. Then Jan Manek, using a new star position from UCAC and a more recent orbit, updated the predicted track from Pisa (Italy) to some km north of Bordeaux (France).
By looking (for field map making) at the Deep Sky Survey, I discovered only a short time before the occultation that the target TYC-1351-00864 is double. The companion, fainter by 2.0 magnitudes, lies 4".55 +/-0".1 apart towards SW (PA = 229 +/-0.5 deg.).
Pic-du-Midi image showing the target star
(47"x47", north up with another field star)

Supposing that this previously unnoticed binarity might have alterated the catalogued UCAC coordinates, and in consequence Manek's update, I wrote an urgent message to [Planoccult] to alert observers against a possible south shift of the path.

Bad weather prevented any observation from Pic-du-Midi and from many parts of s.France and n.Spain or Portugal. Only three misses were recorded from these regions :

- by R.Roy at N44.042 E5.215, not far from Manek's centerline,
- by O.Canales at N41.625 W1.042, not far from Goffin's centerline,
- by R.Goncalves at N39.523 W8.384, who measured on 99 CCD images the closest approach to the blended star+companion at 0".65 +/-0".03 North and 22:24:51.5 +/-4.5 s UTC.

R.Goncalves took complementary CCD exposures of the field some days after, and also improved his astrometric reductions by using the more accurate B1 version of the USNO catalog instead of the old one A2. His revised closest approach is a bit more south than the first one :
0".641 +/-0".034 North and 22:24:51.6 +/-4.5 s UTC.

It seems possible to retrieve the actual track of Deflotte (symbolized by a blue dot in the right figure), providing that we know the true location of the target A relative to the measured photocenter @ of the A+B double star.

The centroid @ depends from the instrumental device (telescope and focal length, focusing accuracy, camera, filters,...), from the atmospheric seeing, and from the centroiding algorithm of the used image processor.
Here I have tried to derive the unknown @-A vector from the measured position of @ relative to the nearby Tycho-2 stars TYC 1351-00220 of magn. 9 and TYC 1351-000938 of magn. 10.5 (data from R. Goncalves, private communication).

My tentative result, at confidence level 95 %, is 0".43 +/-0".23 , PA= 27 +/-33 deg.
(Note that this 0".43 distance is ~10 % of the true A-B separation, that I have measured from Pic-du-Midi independently).

So the corrected appulse of Deflotte to the target A becomes 0".227 +/-0".234 at 22:24:36.8 +/-31 s.

Now we have to compare that with the nominal prediction by Jan Manek for Goncalves' site (Tomar in Portugal) : 0".429 +/-0".080 at 22:24:24 +/-15 s.

Finally the hypothesis of a large south shift of the real path relative to Manek's prediction is supported by the available astrometry, in spite of its modest accuracy (which results for a large part from star binarity).

I suspect a 0".20 +/- 0".23 error toward north (or 5.7 +/-6.7 path widths) of the prediction issued by Jan Manek, caused by an error of comparable importance in the UCAC for this peculiar star.

The figure below shows the possible real path of Deflotte, close to Goffin's initial one, with the associated very large error box.

The Tycho-2 coordinates of the same star might have been affected by less than 0".1, or perhaps not at all..., due to different technology.

I guess that the UCAC catalogue should be discarded from now, for the purpose of update calculations, when the target star appears as an obvious binary by careful inspection of the DSS field.

This provisional conclusion shall be revisited, when more occultations of similar double stars will be identified and measured.

Go back to the 2003 results page is maintained by Eric Frappa -