This paper shows that the chemical makeup of many nearby Milky Way stars can be described by just four shared enrichment patterns. The authors re‑analysed 16 measured element abundances for 199,290 red giant stars with [Fe/H] > -1 and found a low‑dimensional basis that captures the main chemical trends across the disc.

To do this the team re‑projected the 16 abundance measurements into a set of four common patterns. They used that projection as a generative framework — a compact model that describes how mixtures of these patterns reproduce the observed abundances. Then they grouped stars by how much each pattern contributes to a star’s composition and mapped those contributions across the Galaxy.

The results show systematic variations in the pattern mixtures across the disc. Stars that share similar mixtures form coherent ‘‘enrichment pathways’’ that line up with where the stars live, how old they are, and how far they sit above the disc plane. These pathways connect radial growth of the disc with its vertical structure. The authors also find that stars at similar positions along a pathway show matching vertical shifts at different radii, suggesting the low‑dimensional chemical patterns respond to dynamical disturbances in the disc.

The study identifies a change in enrichment behaviour at about 6 Gyr (6 billion years). After that time the disc appears more chemically mixed and shows larger contributions from what the authors call "delayed sources" of enrichment. The paper also finds that the familiar split in alpha (α) abundances — the so‑called α‑bimodality — fits inside the same four‑pattern, low‑dimensional picture, with stars forming continuous sequences of changing pattern contributions.

Important caveats follow from the data and method. The analysis is limited to red giant stars and to stars with [Fe/H] > -1, and it uses 16 measured abundances to define the four patterns. The four‑pattern description is a simplified, low‑dimensional model that highlights major trends but does not by itself identify unique physical causes for each pattern. The timing quoted (about 6 Gyr) is approximate.