U- or V-shaped dip? How to spot the difference?

When searching for exoplanets, the shape of the transit can tell us a lot about what object we could be looking at. For the Planet Hunters NGTS exoplanet transit search, we ask you to identify if a transit is U-shaped or V-shaped, as well as whether there’s stellar variability, data gaps or no significant dip in the flux at all. An exoplanet transiting a star will typically produce a U-shaped dip, but there are situations where that isn’t the case (more on that below). Meanwhile an eclipsing binary (two stars orbiting each other) will produce a V-shaped dip most of the time.

The first plot (Figure 1) shows a clear V-shape produced by an eclipsing binary system. In this case, the transiting star only partially eclipses the target star, meaning that it passes across the edge of the disk of the target star but never passes fully in front. This means that the point of minimum flux doesn’t last long before the flux starts to increase again.

Figure 1: Partial eclipse in an eclipsing binary system

The defining difference between U- and V-shaped dips is the angle of the sides of the transit, or ingress and egress to give them their scientific names. Ingress is when the transit begins and the flux is decreasing to the minimum (position 1 to position 2 in Figure 2), while egress is when the transit is ending and flux starts to increase back to the normal level (position 3 to position 4 in Figure 2).

Figure 2: Diagram of the positions of an object eclipsing a star and the resulting light curve (brightness = flux). [Ref]

V-shaped dips have sides that are at an angle whereas U-shaped transits will have a steeper decrease and increase in the flux, so much so that the sides of the transit will be almost vertical. The reason V-shaped dips have angled sides is because the object blocking out the light is typically (but not always!) another star. The eclipsing star is large (compared to a planet) so takes more time to pass fully in front of the target star, therefore the decrease in flux happens over a significant time period and we get an angled ingress (likewise for the egress as the star stops blocking light).

The angled sides are more pronounced in Figure 1, but don’t be fooled by dips with a curved base like Figure 3 below! If the transit has angled sides then it’s still a V-shape! The curved base of the transit is caused by a phenomenon called ‘limb darkening,’ where the central disk of a star appears brighter than the edge. The eclipsing star in this system is not just grazing the limb of the target star either, which is why the minimum flux of the transit is sustained for a range of phases.

Figure 3: V-shaped eclipsing binary with a curved base

How vertical is vertical? Sadly there isn’t a clear answer to this, which is why we use human vetting rather than just a computer to check these light curves. The example below (Figure 4) is the light curve for confirmed exoplanet HATS-43b, which was classified as U-shaped by all 20 volunteers who viewed it. This is a clear example of the near vertical drop in flux for the sides of the transit. The small radius of the planet compared to its host star means that it almost instantly passes through the ingress and egress phases, compared to the time taken by a larger star in an eclipsing binary system.

Figure 4: HATS-43b. A known exoplanet spotted by Planet Hunters NGTS!

But wait! V-shaped dips can still be exoplanets too! Just like the partial eclipse that produced the sharp, V-shaped dip in Figure 1, an exoplanet can perform a grazing transit where it just crosses the limb of the star and doesn’t go over the centre of its host star’s disk. This will produce a very shallow V-shaped dip, therefore we will get round to searching these classifications for potential exoplanets too! The sides of the dip appear more angled due to the shorter total duration of a grazing transit; it’s very likely that the scale of the x-axis on the plots (the phase) will show a much smaller range of numbers due to how short these transits will be. The ingress and egress times will be similar to a regular transit but the central dip is much shorter. The limb darkening effect also has a more obvious effect on the shape of the dip, which we can see in the light curves below for a near-grazing transit by WASP-174b (Figure 5). The dip has angled sides due to the zoomed in x-axis and has a curved base due to limb darkening. This is a classic curvy V-shape, but it’s also a real exoplanet!

Figure 5: Transit light curves for WASP-174b showing a grazing transit. [Ref]

There isn’t a definitive answer for when a curvy V becomes a regular U-shape, but as always your intuition and best guess is what we want! We hope this blog post makes it easier to spot the differences between U- and V-shaped dips when you’re classifying light curves on the Planet Hunters NGTS site, and remember you can always check the Field Guide or ‘Need some help with this task?’ for more help. There’s also the team of researchers and moderators on the ‘Talk’ forums who will be happy to help!

Sean & the Planet Hunters NGTS Team

About astrosobrien

PhD Student at Queen's University Belfast, searching NGTS data for exoplanets with the help of citizen scientists (ngts.planethunters.org).

7 responses to “U- or V-shaped dip? How to spot the difference?”

  1. George Kountouris says :

    May I send a question, please? When we are observing a V-shaped curve with a curved base (approximate) can we consider that we have to do with the presence of either a rocky planet’s atmosphere or the upper layers of a gas giant? Are there any differences between those 2 cases of light curves?

    • astrosobrien says :

      Hi, the transit light curves will be the same for both cases as far as I know. Certainly for the precision of NGTS we wouldn’t be able to see any difference.

  2. astroly says :

    If a star is larger than a planet, would it actually take less time to pass in front of the star, not more, to produce the V-shaped transit?

    • astrosobrien says :

      It will depend on a few different parameters (i.e. if we choose a fixed period for our example or a fixed orbital separation).

      In the case I was considering (although I could’ve been more explicit), if we assume that the orbiting object (either planet or star) have the same orbital period, then the transit duration of a star will be longer, as the larger radius of the star, compared to a planet, means it has more “distance” that has to pass in front of the host star during the transit ingress and egress.

      If we chose to fix the orbital separation of the system and replaced a planet with a star in our model, then that would decrease the orbital period of the system (Kepler’s 3rd law, the secondary mass has increased), and so it’s possible that the ingress/egress times would be shorter, I think.

Leave a comment