2025ApJ...982...17B

Exoplanet research has moved from the discovery of new classes of planets toward creating a census of exoplanet population demographics across the spectrum of exoplanet and host star masses. This census will constrain possible exoplanet formation and evolution theories. Direct imaging and ground-based microlensing surveys excel at detecting long-period exoplanets. The Roman Space Telescope microlensing survey (RMS) will help complete the census by possibly discovering hundreds of long-period and free-floating exoplanets. Long-period gas giants are difficult to form by the classic core-accretion mechanism, whereas the gas disk gravitational instability (GDGI) mechanism can form gas giants at distances of 10 au and beyond for AFGKM host stars. The Enzo 2.6 adaptive mesh refinement three-dimensional hydrodynamics code is used to model the formation of gas giant protoplanets by GDGI around solar-mass protostars, for future comparison with the RMS, which will provide convincing evidence about whether GDGI is needed to explain exoplanet demographics. Previous Enzo models have shown that a robust GDGI is able to form enough gas giants in a single system to result in the ejection of a significant fraction within ∼2000 yr, a likely source of free-floating exoplanets. The present models investigate the GDGI outcomes for a larger range of initial protoplanetary disk masses and outer disk temperatures than in the previous work of A. P. Boss, resulting in the formation of abundant substellar companions with masses in the range of ∼0.1 to ∼100 M_Jup on orbits ranging from ∼3 to ∼30 au.