This week we have a timely look at a newly (Nov 4) Gaia identified variable in one of our fields, but stay with last weeks theme of discussing caution in the data analysis. The source in question is Gaia23dah, also known ass young stellar object 2MASS J05350274-0500029. The object is also a know X-ray source, and hence clearly a young star. The distance from Gaia is 398pc, so it is clearly a member of the Orion star forming region.

The light curve shown above covers the I-band and R-band data from this and last winter. The gap in-between is the period where the star is too close to the Sun to be observable. In the first part of the data the light curve looks unremarkable flat, with a few outlier data points. While at the start of this winter it looks to behave in the same way, very quickly it looks like it is suddenly slightly fainter but also far more noisy. So what is going on? Well, there are a few points to be made:

  • As one can see in the image displayed on the Gaia alert page for the object, there is a second, slightly brighter (by about 0.5mag) star about 5.4 arcseconds to the North-East of the object. Both stars have almost identical parallaxes, i.e. distances and could be a real physical young, wide binary. The second object is known as a T-Tauri star too. At their distance, the separation indicates that they could be as close as 2100 astronomical units to each other. However, the proper motions are not well matched, so despite them being at the same distance it might just be a chance alignment.
  • The data in this field is very much dominated by observations from one single observatory. The owner of the telescope has changed the camera early on during the current observing season for this field. The old camera had a pixel scale of 2.1 arcseconds, while the new one has much smaller scale of 1.16 arcseconds. Hence, in the older data, the two stars usually merged together into one source, which appeared slightly brighter. Only on very rare, extremely good seeing conditions (sharper images) or when observations by other users with high resolution cameras are added, are the two stars identifiable separately. With the new camera, and the smaller pixels, the two stars appear separate almost all the time.
  • Our light curve extraction software is hence getting confused by this situation and returns all measurements of both stars individually and merged together. If you use the light curve plot option on our server for this object, then you can zoom into details. Indeed you will see that in the “noisy” part of the data for each time there are indeed two brightness measurement. One for the slightly brighter neighbour, and one for the Gaia identified variable.

Thus, again, we need to see how we can improve the software extracting the light curves so that objects like this one (with nearby neighbours) can be analysed reliably. Or we have to except that the analysis of such sources will always be difficult or impossible without detailed manual checking in which images the sources merge into one.