This week we have a short outlook on some general statistical analysis we are working one. As mentioned before, we have used the latest Gaia data release to identify all potential cluster members, and thus young stars, in our survey fields. We have then extracted and calibrated all their HOYS light curves and selected every object that had at least 100 observations in one of the filters. We are in the process of analysing a number of aspects of this sample.
The first thing to note is that we have just over 3000 cluster members in all fields. The lengths of all their light curves now add up to about 20.000yr. Thus, our survey is statistically equivalent to observations of a “typical” young stellar object over twenty thousand years. More importantly, the average time between two data points along this light curve is two days. In any of the broadband filters (V,R,I) the average gap between brightness measurements is one week. Note that the cadence has gotten much smaller in the last few years, as we have a much larger number of images being submitted every day. Furthermore, these ‘averages’ do include the in part several months long gaps in the light curve of some fields when they are unobservable due to the Sun.
One important measure for all the light curves is to identify which of them does represent a variable source. We are using the Stetson index to evaluate this. We show the Stetson index determined for the R-band data of all potential YSOs in the above plot. This measure typically identifies variable stars if the Stetson index is above one. A value of one means that approximately the source varies by the amount of the photometric uncertainties and higher values indicate stronger variability. We can see that a large fraction of the objects is considered as non-variable. But also that there is a continuum of in part very strong variability in our sample. Note that the graph is cut off at a Stetson index of 10, an there are a number of source more variable than this.
We can now investigate just the variable objects and see if they are dippers or bursters (as we have done for all objects as a test in a previous post). We can then count how often e.g. bursts of a given magnitude and duration are present in our data and then evaluate their occurrence rate. Hence, if, for example (note this is a total guess), we find 20 outbursts of one magnitude in our sources, then such outbursts will on average occur every 1000yr in a young star – the most probable rate is a bit different, but this is the main idea…..more at a later point…..