2.7. Requirements for a Successful Monitoring Program

In order to avoid misleading systematic errors of the sort discussed thus far, care must be taken in designing an observational monitoring program. The fairly obvious basic requirements that we should keep in mind are:

1. The monitoring program must have sufficient sampling. Specifically, the observations should be closely spaced in time relative to physical time scales of interest. In terms of total duration, the program should be at least three times as long as the longest time scale to be investigated.

2. The S/N of the light curve must be high enough to detect variations on the shortest interesting physical time scale. Inspection of Fig. 2, for example, demonstrates that if you can measure continuum fluxes to only 10% accuracy, 3 detections of continuum variability in Seyfert 1 galaxies will be extremely rare on time scales as long as months.

3. Systematic effects must be obviated. The data must be as homogeneous as possible. The easiest way to do this is to acquire the data with a single instrument in a stable configuration that is suitable for all the observing conditions expected over the duration of the program.