Mon, 08/05/2019 - 00:37
Recently, a friend of mine told me about one of my stars being listed in Gaia Archive. It happened to be true, so I decided to check some of my future submissions, regarding Gaia variability flag.
It is bothering me, because many of this short period variables are in fact marked as "VARIABLE". Can I still send them to VSX as my discovery or should I make myself co-discoverer or give credits to Gaia?
In uploaded file you can see the star that started it all, NADA-V61.
Far be it for me to speak with the authority of the AAVSO, whose policies on such matters are unknown to me, but as a scientist I was taught that credit for discovery goes to whoever publishes first. So, as far as scientific ethics are concerned, if you discover something on your own, then find that the discovery has been published elsewhere, then while you may take pleasure in your independent discovery, no, you are not the discoverer; not even the co-discover.
I wish the answer was simpler but day after day things get more complicated regarding the plethora of sources flooding the web with data being published in different ways and with different formats, quaility, etc.
We try to keep VSX up to date but 1) we do not want to clutter VSX with poor quality results published automatedly by surveys with very low quality controls and 2) we can't be checking every list published on every journal or survey page.
We recommend using VizieR to check if a star has been published or not in a particular catalogue/list that is not included in VSX. The most popular example nowdays is ATLAS. The list of ATLAS stars requires a lot of pre-processing and can't be imported in a hurry. There are more than 4 million "dubious" stars that might or might not be variable. Those won't be included in VSX.
Most Gaia variables are not in VizieR and a first look at the results published along with the Gaia DR2 catalogue showed also lots of issues, like false alarms, too small number of observations, wrong classifications, formatting. It is unrealistic to require every researcher to look for a specific star in a specific online database every time they are going to submit a star. And it is also unrealistic that we can import and take into account lists that didn't even get to VizieR. We don't have the resources to check everything and have it all covered. And as I mentioned before, we also have to make some decisions regarding what to include and what not. E.g. these dubious ATLAS stars or the stars with less than 80% likelihood of being RR Lyrae stars in the Sesar et al. PS1 list. We have added some lists as not-checked (e.g. a lot of Bochum GDS survey stars) meaning that some o those objects could be duplicates or already be in VSX at slightly different positions, but for several lists, they do not even fill some basic requirements. Some lists do not provide magnitudes (!) or ranges or amplitudes. Do those lists deserve to be taken seriously by a database focusing on star variability? Some lists do not provide coordinates... Do we have to do the work others should have done? You can find all kind of stuff out there.
All I can recommend is that you check VSX and VizieR for new variables. If there is a paper in which the variable is analyzed, that should be considered too (e.g. a SIMBAD or ADS search could be ok), but don't get too crazy about this.
If we import a list that was overdue due to any reasons and that list is worth and has a publication date previous to the publication of your star or anyone else's stars, we will revise the entry changing the primary name to give the proper credit to the group that discovered them first.
Some examples: there are plenty of stars in the Magellanic Clouds or globular clusters being published by surveys without any mention to their extragalactic nature or the fact that they belong to globular clusters.
And the fact is that most of the SMC/LMC variables have already been discovered even decades ago (Harvard variables) or by groups like OGLE and MACHO and globular cluster variables have already been studied and have their own designations in specific papers and catalogues that didn't make it to VSX because VSX is focused on galactic variables (and in principle we also decided to avoid globular clusters due to the ID problems in such crowded environments).
With time, all those survey variables will be given their original designations (e.g. LMC VNNNN or HV NNNN or OGLE-LMC-XXX-NNNN, etc.) and this "contamination" problem will be corrected.
In short, VSX is a living database that is improved every day and every day some mistake may be introduced no matter how careful we are, with the conviction that at the same time, another mistake will be corrected because we spend a lot of time trying to clean up the mess that surrounds us when it comes to data.
So don't overthink it and do not panic. Check VizieR, check VSX and submit your star, if in time it turns out that it had already been found, learn to live with the fact that the credit may be modified, but that will only happen in a few cases. And in any case, the individual analysis you have made (for example, if you did it combining data from several surveys) will always be better than the automated analysis results provided in the lists we import, so your numbers will be kept. In any case your contribution will survive.