Exoplanet Properties by Discovery Method

In class, we’ve discussed how the method we use to discover planets may bias the types of planets we find. For example, larger, more massive planets are easier to find with almost any method of discovering planets. Likely because of this, we have found far more Jupiter-sized exoplanets than Earth-sized planets. I wanted to investigate this further, and examine two planetary properties: the size of the orbit, and the mass of the planet.

To do this, I went to NASA’s exoplanet archive. I downloaded only the columns with the names, method of discovery, orbital distance, and planet’s mass for every discovered exoplanet. Then, I filtered out planets without a listed value for any of these properties. After this, I turned it into a scatter plot with each planet’s orbital distance and mass, all colored based on how it was discovered. The result is below:

A few things are notable right off the bat. First, almost every planet is relatively massive. The absolute smallest planets are only a few times less massive than Earth, and almost every planet is more than .01 – 3x more massive than Earth. But once you pass this threshold, there aren’t huge differences in sensitivity by discovery method. The average mass of planets found by transit is , vs. for exoplanets discovered by radial velocity.

The same isn’t true for distance, though. There is a clear difference even on the plot. The average distance for exoplanets found by transit is only vs. for radial velocity – more than an order of magnitude difference. But both of those are far, far closer than exoplanets found by direct imaging, with an average distance of from their star. In comparison, microlensing – detecting planets by observing their gravity’s influence on photons of light – yields relatively modest numbers. The average distance from planets for microlensing is only .

This shows an important lesson. The planets we see probably aren’t representative of the planets that are actually out there. There are massive variations in the kinds of planets we find just based on the method we use to discover them. And there’s very good reason to think that all of them are biased in favor of finding large planets, meaning our data is likely extremely unrepresentative of the true distribution.