The star for this week is situated in the Pelican Nebula (IC5070) target field. But it is a background object with a distance of about 1650pc. However, it is a variable – specifically an eclipsing binary. It is relatively faint, about V=17.7mag on average. thus the photometry is a bit noisy.
The plots above show the data for the V, R, and I-band filters. In the top right panel we just display the light curve. The top left panel shows the periodogram. At the bottom we have two cycles of the phase folded light curve. These show that the two eclipses (primary and secondary) are basically the same depth and they are roughly the same depth in all filters, about 0.2mag. They are harder to see in the shorter wavelengths filters as the source is much fainter, and the data are hence more noisy.
In the periodogram one can see that the highest peak is at just over one day, and not at the actual period shown in the plot. This is a common problem with sources like this. Each period has two identical dimming events and thus usually half the real period is found. Note we find a period of 2.35234116d (2d 8h 27m 22.3s). In the literature a number of other periods are listed: 2.3518976d and 2.353d. The differences seem small, -38.3s and +56.9s, respectively. However, if one phase folds the data with either of the literature periods, the dimming events do not line up and the phase folded light curve looks just like noise. This is caused by the very long dataset we have (over 8yr). Thus, we can determine the period very accurately.
One can estimate the accuracy by assuming that if there is a shift of more than 10% from the first to the last period that is phase folded, then one could not see the dimming events clearly anymore. Given we have about 1000 cycles in the data, the accuracy is hence one 10,000th of the period. This is about 2s.