Presentation #132.06 in the session Extrasolar Planets: Populations I.
The relationship between exoplanet mass and radius is an important input for planet formation and evolution studies. The existence of a range of masses where rocky, water-rich, and volatile-rich planets can have similar masses but differing radii makes the mass-radius relationship complicated. The relatively small number of planets with both measured mass via radial velocities (RVs) and measured radius via transits limits our understanding of the mass-radius relationship in this mass range.
An important feature of a mass-radius relation is the relative proportion of planets categorized by composition, for example rocky and volatile as in Otegi . This proportion can be difficult to determine when planets in different categories have similar masses but different radii. We present a new method of constraining that relative proportion combining independent RV and transit data. We jointly analyze the well-characterized HARPS+CORALIE and CKS RV surveys combined with the Kepler transit survey for planets with mass between 2 and 50 MEarth and periods between 2 and 200 days. By not requiring that planets have both measured masses and radii, this data provides a large sample size with which we can perform demographic analysis. We investigate the rocky vs. volatile categories in the Otegi mass-radius relation , and generalize to several composition categories such as rocky, rock + H2 envelope, water-rich, and volatile-rich. We extend the inference methods described in  by using likelihood-based methods that allow model comparison. We find that simple abundance power laws in mass and period are sufficient to replicate the complex period-radius structure near the radius valley. Similar to , we find that, in the mass range where rocky and volatile-rich planets overlap (~5 to ~25 MEarth), about 60% of the planets can be considered rocky or rocky with a significant Hydrogen envelope. We discuss how that proportion varies with planet mass and radius, which informs the characterization of planets for which only radius is known.
 J. F. Otegi, F. Bouchy and R. Helled 2020, Astronomy & Astrophysics 634, A43
 M Kunimoto and S Bryson 2021, The Astronomical Journal 161 (2), 69