4 Instrumental effects

Systematic photometric effects in FGS archival data were studied by Zwintz et al. (1999) as part of an analysis of the northern Hubble Deep Field program. Not surprisingly, brightness variations modulated by the HST orbit frequency, or in resonance to it, were detected. These orbit-induced variations are caused by scattered light from the illuminated Earth visible from the spacecraft. This occurs in about 15% of the high quality data sets with noise levels below 100 ppm. A further significant source for increasing noise levels are the wings of the South Atlantic Anomaly (SAA). During the passage through the actual SAA, however, the FGS instruments are switched off to protect the sensitive PMTs. A full correction of the FGS photometry for stray light effects will probably shift most of the guide star data sets classified as "o'' to one of the other three variability groups.

Compared to the stray light contamination, the nature of "t'' group variables cannot be unambiguously interpreted yet. For example, some of the data sets reveal strong intensity offsets between consecutive orbits. Other data sets show abrupt drops in intensity, which turn out to be limited to the X or Y channel of the FGSs (Fig. 2) and which last for one or more orbits with occasional rise to the previous value (Zwintz et al. 1999). In some cases we have "saved'' these observations by eliminating the faulty channel, but taking an increase of the noise level into account.

While a detailed analysis of the source of these effects remains to be undertaken, it appears likely that they are due to double S-curves, leading to slightly different fine lock positions and correspondingly different photometric sensitivity. It is known that in certain areas of FGS2 such double S-curves occur. They may also be caused by unresolved binaries; depending on their position angle, they can influence either or both channels of the FGS in question.

We do not discuss instrumental effects caused by different sensitivity, dead time correction and color transformation of the FGS instruments involved in our study but refer to Weiss et al. (1999) and to a paper in preparation on probably constant Guide Stars.

Figure 2: GS0916600446, an example for the "dropping channel'' problem

   

Table 3: Strömgren photometry obtained at ESO by S.F.: variable guide star number (VGS); Guide Star Catalog number (GSC); Johnson V magnitude; Strömgren b-y, m1 and c1 values

VGS	GS	V	b-y	m1	c1
11	0495601167	11.62	0.321	0.119	0.411
12	0502500568	9.80	0.219	0.173	0.932
13	0685403208	11.46	0.310	0.104	0.784
14	0685601078	11.28	0.119	0.065	0.636
15	0739300128	9.97	0.229	-0.017	0.884
17	0742602146	11.66	0.206	0.164	1.017
18	0779501427	8.89	0.965	0.903	0.201
19	0781801912	11.15	0.179	0.243	0.751
20	0825203215	9.54	0.445	0.220	0.668