The Milky Way's formation involved distinct phases resulting in its halo and disk components. Understanding its assembly requires precise ages and abundances of numerous stars, extending to the oldest possible ages (around 14 Gyr). Subgiant stars offer accurate age determination due to their luminosity's direct dependence on age, and their chemical composition is readily determined from their spectra. However, their shorter lifespan makes them rare, necessitating large-scale surveys. This study leverages the Gaia eDR3 and LAMOST DR7 datasets to overcome this limitation, identifying approximately 250,000 subgiant stars. The goal is to reconstruct the Milky Way's formation and enrichment history, focusing particularly on its early phases. The research uses subgiants as tracers to gain insights into the timing and processes involved in the formation of the Galaxy's various components, bridging the gap between theoretical models and observational data.

Previous research has qualitatively described the Milky Way's formation phases but lacked the sample size and precision for a definitive quantitative understanding. Earlier studies hinted at features in the age-metallicity distribution but couldn't definitively characterize the Galactic formation history due to limitations in data size and accuracy. While studies exist on the chemical composition of the Milky Way and its disk, the timing of events and the processes behind the various stellar populations were not fully resolved. This study builds upon previous work by providing a much larger and more precise dataset, which allows for a more detailed and nuanced interpretation of the Milky Way's assembly history.

The study utilizes data from the Gaia mission's eDR3 release and the LAMOST spectroscopic survey's DR7 release to identify a sample of approximately 250,000 subgiant stars based on their position in the effective temperature (Teff)-absolute magnitude (MK) diagram. Stellar ages (τ) were estimated by fitting the data to Yonsei-Yale (YY) stellar isochrones using a Bayesian approach, incorporating astrometric distances (parallaxes), apparent magnitudes (fluxes), and spectroscopic chemical abundances ([Fe/H]). The age uncertainty across the sample (1.5 Gyr to 13.3 Gyr) is a median of only 7.5%. The lower age limit is due to the difficulty of distinguishing younger subgiants from other evolutionary phases. This large sample, covering a wide spatial volume and range in age and metallicity, allows for the investigation of the Milky Way's assembly history. The researchers then analyzed the resulting stellar age-metallicity distribution, p(τ, [Fe/H]), to reveal the chemical enrichment history and the timing of different stellar populations' formation. The sample was further divided into subsamples using stellar angular momentum (J) and α-enhancement ([α/Fe]) to separate stars based on their orbital properties and chemical composition. This refined analysis aided in identifying distinct sequences within the age-metallicity distribution, which were crucial for uncovering details of the formation history, particularly the early stages. Finally, the volume-corrected two-dimensional distribution p(r, [Fe/H]) was analyzed by separating stars based on their orbital properties (retrograde versus prograde) to further enhance understanding of the merger events that shaped the Milky Way.

The analysis of the stellar age-metallicity distribution revealed two distinct sequences separated at an age of approximately 6 Gyr. The younger sequence (τ < 6 Gyr) represents the late phase of dynamically quiescent Galactic disk formation with significant stellar radial migration. The older sequence (τ > 6 Gyr) represents the earlier formation of the stellar halo and the old α-enhanced (thick) disk. The formation of the old disk started about 13 Gyr ago (0.8 Gyr after the Big Bang), extending over 5-6 Gyr. The interstellar medium (ISM) was continuously enriched by more than 1 dex in [Fe/H] during this period, indicating thorough spatial mixing within the ISM. A majority of stars in this old disk formed around 11 Gyr ago, coinciding with the merger of the Gaia-Sausage-Enceladus satellite galaxy. This merger enhanced star formation in the old disk, as indicated by a peak in the star formation rate at approximately 11.2 Gyr ago. The merger between the old disk and the Gaia-Sausage-Enceladus satellite galaxy was largely complete 11 Gyr ago. The study also revealed a Z-shaped structure in the age-metallicity distribution, further highlighting the complex interplay between different stellar populations.

The findings significantly advance our understanding of the Milky Way's early formation and enrichment history. The clear age-metallicity relation and its tightness indicate efficient mixing of the ISM throughout the old disk's formation. The coincidence of the star formation peak with the Gaia-Sausage-Enceladus merger suggests a causal relationship. The results highlight the importance of major merger events in shaping galaxy evolution. The study also demonstrated the power of combining precise stellar ages and abundances with kinematic information to decipher complex galactic assembly processes. This approach provides a detailed, time-resolved picture of the Milky Way's formation, complementing theoretical models and providing crucial observational constraints.

This study provides a detailed, time-resolved picture of the Milky Way's early formation history using a large sample of subgiant stars. The findings reveal distinct formation phases, with the old disk starting to form 13 Gyr ago, significantly overlapping with halo formation. The Gaia-Sausage-Enceladus merger played a crucial role in shaping the old disk’s star formation. Future research could focus on expanding the sample size further, incorporating more detailed chemical abundance data, and applying similar methods to other galaxies to test the universality of these findings.

While the sample size is substantial, it may not be fully representative of the entire Milky Way. The reliance on specific stellar evolutionary models could introduce some uncertainties. Additionally, the analysis focuses primarily on the inner regions of the Galaxy; applying the same methodology to outer regions would broaden the study’s scope. Potential biases in the sample selection due to observational limitations or the specific choice of isochrones used for age estimation should be considered.