The Architectures of Planetary Systems: Population Synthesis Meets Observations

Advancing our understanding of planet formation is a prime motivation for the search for exoplanets. While it is now widely recognized that multiple planets per system are common, their mutual relationships are still largely unexplored. This thesis investigates such relationships by confronting simulated planet populations with observed planetary systems. To draw conclusions about the formation environments of planets that we observe today, the causal connections between their properties and those of their natal protoplanetary disks must first be established. In a data-driven approach, I identify the most predictive initial conditions of a planet formation model and show that N-body interactions affect primarily low-mass planets. These insights are then used to study the relations between super-Earths on short orbits and outer giant planets. I find a connection between the com- position of simulated planets and the architectures of their systems. This gives rise to the testable hypothesis that high-density inner super-Earths point to a giant companion in the same system. The analysis also suggests that dynamically active giant planets frequently destroy systems of inner super-Earths. This is compatible with the discovery of one of the most eccentric warm Jupiters known that I present in this thesis. I demonstrate in a tidal evolution analysis that this planet is not the progenitor of a hot Jupiter during its high-eccentricity migration. To explore variations of these trends as a function of stellar host mass, I confront the CARMENES M dwarf survey with a synthetic population of planets around low-mass stars. A striking discrepancy is the observed existence of giant planets around very low- mass stars, which can not be reproduced by our model. Future planet formation theories must explain also this peculiar finding