One of the issues with our quite in-homogeneous data is that it can contain photometric outliers. In other words, these are brightness measurements which appear to have a small uncertainty but do not agree with other measurements taken with other telescopes at roughly the same time. This can be caused by a variety of reasons, such as cosmic ray hits, problems with the astrometric solution in the image, issues with the calibration, or a nearby neighbouring star. Thus, we need to have a routine available that can identify potentially erroneous data points in a light curve. It is important to exclude them from the analysis, as they can alter the results.
However, this is not a trivial task. For non-variable stars, looking for deviations from the mean magnitude is simple. But this does not work for variables – which we are interested in. A student has written a first procedure that looks for ‘outlier’ photometry in the light curves as well as in colour magnitude plots. The results for an example source are shown in the above light curve. All photometry points that get identified as potentially problematic are enclosed with a light blue border. While this works quite well for most of the data, you can see that for dates between 60300 and 60400 the object seems to have a real dimming event – all measurements are fainter in all filters (with some outliers). However, all these points are flagged as outliers. Thus, we need to be a bit more clever about this and consider changes in all filters simultaneously.
Hence, we have made this a final year undergraduate project, and so far there are some promising first results – for this object the new procedure leaves in the dimming event points! – and we’ll wait for the project to finish before writing another post with the final results for this source to show the comparison. One thing is clear: The more data we have for the brightness measurements in all filters, the simpler it is to identify outliers. Hence the goal of our project to observe every field in all filters at least once a day….please keep observing 🙂
PS: The database should re-open soon for image submission!