Duplicate entries

Hello everybody,

I'd like to start out this forum with a good example of what we do every
day to improve VSX.

Personally, I devote a great deal of my time to hiding duplicate records
in our database; we don't delete records in VSX but we hide them from
view once we find them redundant or wrong.  As an example, a star
already included in the GCVS may have been observed by the ASAS survey
and ended up in its variable star catalogue too, thus we will have two
entries for the same star in VSX with the information included in each
source. Since the ASAS resolution is not so good, sometimes a position
only is not enough to be sure a cross-identification can be made. So we
have cross-identifications already coming from the ASAS catalogue and
new cross-identifications made by ourselves.

Every day we hide several of these duplicate records, creating a
"primary record" with all the available information. If ASAS
observations are useful to improve the results in the GCVS record (the
GCVS record will always be the one chosen to be the primary record and
be revised since it has the official IAU designation of the variable),
they are used to do that along with data from all other available
surveys or catalogues that can improve the information we offer. This
example is valid for all kind of duplicates that we can find in the
database (e.g. NSVS+ASAS, GCVS+NSVS, GCVS+OGLE, CRTS+ASAS and so on and
so forth).

But there are also more weird things we have to deal with! Sometimes
automated analysis of the survey results leads to the inclusion of
duplicates within the same catalogue. A common example is a bright red
variable completely saturated in a red-sensitive survey that is
contaminating fainter nearby stars causing them to appear variable (and
with the same period, of course!). You may even find that the target
(usually an already known variable because it is bright) is not included
in that catalogue (because it is above its saturation limit) but all the
contaminated neighbours (actually non-variable stars) are. This happens
a lot in crowded fields so the OGLE catalogues are prone to this
problem.

The example I want to share is the duplicate record-breaking entry I had
to fix. I did it on 2011, May 2. The star in question was NSV 22788:

http://www.aavso.org/vsx/index.php?view=detail.top&oid=61223

It varies between 11.5 and 16.5 in R so it is much brighter (thus
saturated) in OGLE Ic images. The actual variable is not included in the
OGLE catalogue but it turns out that the following 32 (!!) records in
the OGLE list published in 2005A&A...443..143G were actually faint
objects all around the variable star's position and contaminated by its
light:

OGLE BUL-SC43 V1149
OGLE BUL-SC43 V1153
OGLE BUL-SC43 V1143
OGLE BUL-SC43 V1154
OGLE BUL-SC43 V1156
OGLE BUL-SC43 V1157
OGLE BUL-SC43 V1159
OGLE BUL-SC43 V1168
OGLE BUL-SC43 V1170
OGLE BUL-SC43 V1171
OGLE BUL-SC43 V1172
OGLE BUL-SC43 V1174
OGLE BUL-SC43 V1175
OGLE BUL-SC43 V1176
OGLE BUL-SC43 V1177
OGLE BUL-SC43 V1178
OGLE BUL-SC43 V1215
OGLE BUL-SC43 V1216
OGLE BUL-SC43 V1217
OGLE BUL-SC43 V1246
OGLE BUL-SC43 V1247
OGLE BUL-SC43 V1255
OGLE BUL-SC43 V1257
OGLE BUL-SC43 V1258
OGLE BUL-SC43 V1273
OGLE BUL-SC43 V1274
OGLE BUL-SC43 V1275
OGLE BUL-SC43 V1279
OGLE BUL-SC43 V1281
OGLE BUL-SC43 V1282
OGLE BUL-SC43 V1283
OGLE BUL-SC43 V1286

Amazing, isn't it? Fortunately that was an extreme case but there are
still some to solve (no more than 2 duplicates now, all cases with
several objects werer already addressed).

This is a good example of what we do to clean up VSX and I wanted to
share it with you.

If you find a duplicate entry in VSX, let us know and we will fix it.

Cheers,
Sebastian