Error Reduction Strategy

The strategy for reducing the error in the proper motion results is to choose among several conflicting constraints in order to reduce the error in the rate of relative distance change between pairs of quasars.

Choose the highest rate of distance change.
Choose longer distances. The position of a star has a fixed uncertainty, therefore the longer the distance, the smaller the error.
Choose distances short enough to fit in one CCD frame. Astrometric precision degrades significantly if multiple CCD frames must be merged together. If the CCD has a small field of view try positioning a quasar pair on opposite corners of the chip. (For square CCDs, the diagonal is 41 percent longer than the edge)
Since very small changes in pixel distances must be detected, temperature regulation is important in order to reduce error contribution from thermal expansion of the CCD chip.
A CCD autoguider will greatly improve the accuracy of the results : SBIG has recently announced the ST-7 and ST-8 self-guiding cameras with two CCD detectors, one for guiding the other for collecting the image. (Oct 95 Sky and Telescope p.78)
Given the long term nature of the project, it is advisable to make annual exposures when the quasars reach their maximum elevation, for the quasars in the cancer constellation this means wintertime. However if atmospheric seeing improves in the early morning, the exposures may have to be taken earlier in the year.
The advantage of a smaller telescope is that the seeing is less of a problem as Levy pointed out during the comet encounter with Jupiter, views of collision sites looked sharper from the smaller telescopes.
Observations should be made at the same sky elevation to avoid atmospheric refraction effects.
Observations should be made with the same telescope and CCD. However only the angles or the ratio between adjacent sides of a rectangle are monitored then a wide varied of data from many sources could be used.

CCD versus Photographic Astrometry

In the past, using film meant very long exposures for 18th mag stars, in which cumulative tracking errors and oscillating seeing conditions greatly degraded the star image. Unless one worked with expensive and difficult glass plates, the spatial distortions due to warping during exposure, shrinking during processing or temperature variations etc... lead to great problems when it comes time to scan it into the computer.

The advent of CCD imagers, with their fixed pixel positions, makes quantitative stellar position measurements very accurate. Their large quantum efficiency reduces by a factor of 10 the required exposure time for any fixed aperture, or for the same exposure time, a 16 inch telescope would act like a 1 meter monster telescope.

The star images will be blurred, however the first step is to take a sufficiently large number of CCD exposures and rejecting them when the seeing is bad; either during the same night or on subsequent nights. With the inexpensive nature of the storage medium, it now becomes practical to be very picky and choose say only one out of a dozen CCD exposures, this would otherwise become prohibitively expensive with glass plate film.

Then by careful position reduction using accurate point-spread function fitting procedures in software packages like Astrometrica, CCDAST, DAOPHOT etc... Find the centroid of the star at sub-pixel accuracy, you should be able to follow the proper motion by using one clear exposure per month over the span of several years. With a sufficient number of points, the random errors due to seeing and low resolution should in principle be averaged out and the only remaining trend would be the systematic proper motion.

Back to Amateur Astrometry.