学术文献库

Understanding planet formation requires robust population studies, which are designed to reveal trends in planet properties. In this work, we aim to determine if different methods for selecting populations of exoplanets for atmospheric characterization with JWST could influence population-level inferences. We generate three hypothetical surveys of super-Earths/sub-Neptunes, each spanning a similar radius-insolation flux space. The survey samples are constructed based on three different selection criteria (evenly-spaced-by-eye, binned, and a quantitative selection function). Using an injection-recovery technique, we test how robustly individual-planet atmospheric parameters and population-level parameters can be retrieved. We find that all three survey designs result in equally suitable targets for individual atmospheric characterization, but not equally suitable targets for constraining population parameters. Only samples constructed with a quantitative method or that are sufficiently evenly-spaced-by-eye result in robust population parameter constraints. Furthermore, we find that the sample with the best targets for individual atmospheric study does not necessarily result in the best constrained population parameters. The method of sample selection must be considered. We also find that there may be large variability in population-level results with a sample that is small enough to fit in a single JWST cycle ($\sim$12 planets), suggesting that the most successful population-level analyses will be multi-cycle. Lastly, we infer that our exploration of sample selection is limited by the small number of transiting planets with measured masses around bright stars. Our results can guide future development of programs that aim to determine underlying trends in exoplanet atmospheric properties and, by extension, formation and evolution processes.