We briefly describe the effects which limit the general use of the method. In so doing, we also indicate the process one can use in deciding if there is a problem, and estimating the chances of substantial improvement.
The FWHM of all sources in the field should
be 7
(after smoothing with a 3
Gaussian). If any source is smaller than this value, it is likely
that aspect problems are minimal and little is to be gained by
applying the dewobbling method.
If there is only a single source in the field, without a priori knowledge or further analysis it is difficult to determine whether a distribution significantly larger than the ideal PRF is caused by source structure or aspect smearing. The best approach in this case is to examine the image for each OBI separately to see if some or all are smaller than the total image (i.e. OBI aspect solutions are different).
It is important that the phase of the wobble is maintained. This is
ensured if there is no "reset" of the space craft clock during an
observation. If an observation has a begin and end time/date that
includes a reset, it will be necessary to divide the data into two
segments with a time filter before proceeding to the main analysis.
Dates of clock resets (Table 1) are provided by MPE:
http://www.ROSAT.mpe-garching.mpg.de/prp/timcor.html.
In most cases, the reference source (i.e. the source used for centroiding) will be the same as the target source, but this is not required. Ideally, the reference source should be unresolved in the absence of aspect errors and it should not be embedded in high brightness diffuse emission (e.g. the core of M 87 does not work because of the bright emission from the Virgo Cluster gas). Both of these considerations are important for the operation of the centroiding algorithm, but neither is an absolute imperative. For accurate centroiding, the reference source needs to stand well above any extended component.
Obviously the prime concern is that there be enough counts in a phase bin to successfully measure the centroid. The last item is usually the determining factor, and as a rule of thumb, it is possible to use 10 phase bins on a source of 0.1 counts/s. We have tested a strong source to see the effect of increasing the number of phase bins. In Fig. 1, we show the results of several runs on an observation of HZ 43 (12 counts/s). This figure demonstrates that ten phase bins is a reasonable choice, but that there is little to be gained by using more than 20 phase bins.
![]() |
Figure 1: The FWHM of a HZ43 (observation number rh142545) observation was measured for multiple dewobble runs while increasing the number of phase bins |
Copyright The European Southern Observatory (ESO)