In essence, the variations and irregularities are caused by "shock transits". We should be aware of them when analyzing light curves. I believe that many of the exoplanet candidates in Kepler's data release have bow shocks because most of them lie very close to their respective host stars.
Also, I think that bow shocks are bright, especially when the planet plows through material from a massive coronal mass ejection (CME). Hence some bow shock transits should cause a spike (instead of a dip) in the light curve. This spike, I assume, should occur a short moment before the planet crosses the star, or shortly before the "ingress".
However, there's a disclaimer I'd like to add to this post: The paper says that the effect of the shock transit shows up in the near-UV light curves. But Kepler's light curves are photometric optical light curves, so i am making an inference that effects of bow shocks might show up in Kepler's optical light curves as well. (I will dig more about this topic and post as soon as i get more info).
But For now, take a look at the snapshot below and note the pattern of how the bow shock's transit shows itself in the light curves. And more importantly, a reversed pattern might show up for planets that have "behind shocks". The main principle is that bow shocks can affect light curves.
Good luck! And don't be shocked if you find yourself one of the co-discoverers of a planet with an amazing bow shock!
