3. X-ray light curves
The number of source counts accumulated during the 1.7 days of the ROSAT survey observation is too small to allow the construction of a statistically significant survey X-ray light curve. Therefore, we restrict the study of the X-ray variability of Ton S180 to the pointed observations in 1992 and 1993.
In order to prevent that the X-ray image is shaded accidentally by the wires of the PSPC entrance window supporting grids, the telescope axis is wobbled with a period of 400s and an amplitude of 3' (Briel, 1991), while the telescope is pointed at a target source. This causes the source's image to move across the wires, averaging out the obscuration effects. In reconstructing a stable X-ray image the wobble is removed a posteriori by applying the attitude solution. As the pointing of the telescope is actively controlled by the Attitude Measurement and Control System of the satellite, an attitude jitter is superimposed on the linear wobble movement. Consequently, the point source image performs a random walk over the detector along the wobble path, which makes the actual shading of the image by the supporting grids unpredictable. A reliable intensity estimation of the point source image can therefore only be obtained by averaging the count rate over an entire wobble period. Nevertheless, different wobble pathes cause different averaged count rates, but the variation of the flux determination in the wobble mode is less than 4% (Hasinger, private communication). As a consequence of the wobble mode, the smallest time scale reproducibly measurable with the PSPC in pointed observations is 400s.
The QSO was observed at four epochs, each separated by about half a year. As can be seen from Table 1, the pointed observations in June 1992 comprise six orbits which are grouped to three pairs of adjacent orbits separated by one and by eight and a half days, respectively. The second observational epoch in Dec 1992 consists of two adjacent orbits. The first observation in 1993 lasting only one orbit took place 23 days later. In June 1993 the QSO was observed during six successive orbits.
The intra-orbit variability of the source is determined by splitting up the total exposure per orbit into 400s bins, starting with the arrival time of the first incident photon. Usually, the exposure per orbit is not an integer multiple of 400s. The incomplete last bin has to be rejected for the purpose of contructing a light curve, because the averaging over an uncomplete wobble period would yield unreliable count rates. For the count rate per 400s bin all vignetting corrected counts within the source extraction area are gathered which are registered within the time limits of the bin considered. The source count rate per bin is obtained by subtracting the extrapolated background count rate referring to the source extraction area. Finally, a dead time correction has to be applied which is mainly determined by the master veto anti-coincidence rate.
All of the observations in 1992 contain three 400s bins, whereas the 1993 observations are shorter with the exception of that one in January and of the last pointing in June which contain four 400s bins. The intra-orbit light curves of those observations comprising three or more 400s bins are shown in Fig. 3. Obviously, the count rate varies smoothly over the orbits. There are no "noisy" orbits such as those found in the case of the BL Lac object PKS 2155-304 (Brinkmann et al., 1994). The maximum intra-orbit variation occurs in the fifth orbit of the June 1992 observation. There the count rate varies by 18% over three bins corresponding to a formal doubling time of 0.08 days.
The orbit-to-orbit X-ray light curves of Ton S180 are given in the upper panel of Fig. 4. The data points are obtained by summing up the count rates of the 400s bins and by dividing the sum by the integer multiple of 400s contained in the orbit exposure.
In June, 1992, the (0.1 - 2.4)keV count rate rises from a level of 3.34 cts/s in the first two orbits to 6.75 cts/s almost one day later. The upper limit of the corresponding doubling time is 0.99 days. In the following orbit the count rate drops by 13% corresponding to a negative doubling time of 0.52 days. Eight and a half days later the source is found to emit at a mean count rate level of 5.2 cts/s. The undersampling of the light curve does not allow to state definitely that the decay time of the outburst in orbit No.3 is much longer (days to weeks) as the rise time (1 day).
During the two orbits in Dec 1992 the source emits at a count rate level of about 3.6 cts/s and does not vary significantly. 23 days later in Jan, 1993, the source is found at a slightly higher count rate level of 4 cts/s, but, again, the intra-orbit variations are negligible.
During the observation in mid 1993 the count rate drops from a level, which equals almost that one the 1992 June observation started with, by 31% over three orbits and rises again to the initial level in three further orbits. The formal doubling time of this variation is 0.64 days.
Finally it should be noted that the mean count rate of Ton S180 during the survey observation, 1.62 cts/s, is lower by a factor of two than the mean rate during the pointed observations. Therefore, the amplitude of the long-term variability of Ton S180 is quite as large as that one of the orbit-to-orbit variations with a time scale of one and a half hour.
In the lower panel of Fig. 4 the hardness ratios and that one of the entire carbon window of the PSPC's transmission, , are shown. If we exclude orbit-to-orbit variations of the quasar's spectrum as unrealistic, there are no signs of remarkable spectral variability in the June observations in 1992 and 1993, respectively. On the other hand, the hardness ratios measured in Dec 1992, , and in Jan 1993, , have inconsistent values, even if we adopt appreciably higher uncertainties, possibly indicated by the putative variability of adjacent orbits, as the pure statistical errors given.
© European Southern Observatory (ESO) 1997