Quality controls
We have performed data checks and quality controls on the data stored in the
database, both on internal and external
bases. Many inconsistencies
were found. When an obvious solution could be found, corrections were applied
to the stellar identifications or photometric data.
If not, the number of measurements has been set to null (0), so that the
suspect observations do not influence the computation of mean values,
when relevant.
We have prepared lists of errors and problems found. They can be reached
from the description of the characteristics of
each photometric system.
Each time there is at least two independent sources of measurements for a
star in a photometric system, we have compared them and looked for
larger than normal differences.
When some were found, we have carefully checked the values
entered in the computer or the published identifications.
However, in many cases, it was not possible to find any reliable explanation for the discrepancies.
The problems may be due either to the observation of a wrong object
or to some misprint.
We have also compared similar data from several photometric systems.
For example, we have checked that the V magnitudes published for the
same star in various systems are in good agreement. Again, any abnormal
difference may reveal a problem on the measurements or the star identification.
We shall also work graphically. We plot diagrams with photometric indices of similar
astrophysical content (for example B-V versus b-y) and examine the most deviant stars.
Back to the GCPD front-page
epaunzen@physics.muni.cz
Last update: 16 March 1996