ShangGu-2: Heated Metal-Poor Disc Substructure

• ShangGu-2 is a chemo-dynamically identified substructure composed of metal-poor disc stars that have been kinematically heated, showing disc-like rotation with high vertical motion.
• It was isolated using HDBSCAN clustering in 4D dynamical space from LAMOST DR9 stars cross-matched with Gaia DR3 astrometry and StarHorse distances.
• Its dual [Mg/Fe] sequences and distinct kinematic signature differentiate it from accreted populations, supporting an interpretation as a heated, in-situ metal-poor disc component.

Searching arXiv for the cited paper and any directly relevant related work on the metal-poor Galactic disc and identified substructures. ShangGu-2 is a newly identified dynamical substructure within the Galactic metal-poor disc, isolated in a chemo-dynamical analysis of a LAMOST-based sample of stars with $[\mathrm{Fe/H}] < -1.0$. In the naming scheme of "Chasing the formation history of the Galactic metal-poor disc," it corresponds to Cluster 5 among the 28 dynamically unclassified DTGs and is interpreted primarily as a heated disc component rather than an accreted remnant. The defining interpretation is explicit: ShangGu-2 “is composed of stars from the Milky Way’s metal-poor disc, which have been kinematically heated recently, thereby acquiring significant vertical motions” (Hou et al., 21 Jul 2025).

1. Discovery context and nomenclature

ShangGu-2 emerged only after a two-stage decomposition of the metal-poor stellar sample. The study first separated stars into kinematically defined disc and merger-remnant components, and then applied clustering in dynamical space to identify coherent substructures among the residual populations. Within that framework, ShangGu-2 is not a primary sample defined by a single threshold in chemistry or kinematics; rather, it is a clustered overdensity in phase space that was identified after the coarse classification step.

The relevant naming scheme is internal to the study. Among the dynamically tagged groups, or DTGs, that were not matched to the named structures—cold disc, hot disc, GSE, Pontus, Thamnos, Sequoia, Helmi Streams, and Wukong—the authors grouped three residual clusters into newly designated substructures: ShangGu-1, ShangGu-2, and ShangGu-3. ShangGu-2 is the middle member of this trio and is the one interpreted most directly as a heated metal-poor disc population (Hou et al., 21 Jul 2025).

2. Data set and identification procedure

The parent sample is the LAMOST DR9 metal-poor catalog, originally comprising about 100,000 stars with $[\mathrm{Fe/H}] < -1.0$. For the kinematic analysis, this catalog was cross-matched with Gaia DR3 astrometry, StarHorse photo-astrometric distances, and LAMOST DR10 or Gaia DR3 radial velocities. The sample was filtered by removing stars with relative distance uncertainty greater than 20\% and stars with RUWE $> 1.4$; when multiple LAMOST observations existed, only the highest-S/N spectrum was retained.

The study computed positions, velocities, integrals of motion, and orbital parameters with Agama in the McMillan (2017) Milky Way potential, propagating uncertainties through 10,000 realizations per star. The main dynamical analysis then used a cleaner subsample of 46,178 stars with reliable kinematics. A separate age analysis cross-matched this sample with the Xiang et al. ages, yielding 2,517 subgiants with reliable age estimates.

ShangGu-2 was isolated through HDBSCAN clustering in the 4D dynamical space $(E, J_\phi, J_r, J_z)$ with min_cluster_size = 15, min_samples = 6, metric = euclidean, and cluster selection method = leaf. This procedure produced 6,549 stars in 222 DTGs; 28 DTGs were not matched to known structures and were then grouped into six residual clusters, one of which is ShangGu-2 (Hou et al., 21 Jul 2025).

3. Kinematic signature

The defining kinematic property of ShangGu-2 is the combination of disc-like azimuthal rotation with anomalously large vertical motion. Its stars have $V_\phi \lesssim -200\ \mathrm{km\,s^{-1}}$, which aligns with the rotational characteristics of the kinematically cold disc, but they also have $|V_z| \gtrsim 100\ \mathrm{km\,s^{-1}}$, making them vertically much hotter than a conventional cold-disc population.

This combination is central to the cluster’s interpretation. In the study’s comparative kinematic diagrams, ShangGu-2 resembles the cold disc in $V_\phi$ while differing sharply in $V_z$. The result is a population that is rotationally disc-like but dynamically disturbed in the vertical direction. That pattern is consistent with kinematic heating, and the paper treats it as the primary observational basis for distinguishing ShangGu-2 from both ordinary cold-disc stars and more clearly accreted populations (Hou et al., 21 Jul 2025).

4. Chemical abundance pattern and age information

Chemically, ShangGu-2 spans a metallicity range from $[\mathrm{Fe/H}] = -2.27$ to $-1.13$, with mean $[\mathrm{Fe/H}] < -1.0$0. Its most important abundance diagnostic is the structure it traces in the $[\mathrm{Fe/H}] < -1.0$1–$[\mathrm{Fe/H}] < -1.0$2 plane: the cluster shows two distinct sequences, “just like our kinematically cold disc sample.” The paper does not report a unique $[\mathrm{Fe/H}] < -1.0$3-knee for ShangGu-2 itself, nor a separate fitted $[\mathrm{Fe/H}] < -1.0$4-$[\mathrm{Fe/H}] < -1.0$5 relation dedicated solely to this cluster. The chemical argument is therefore comparative rather than parametric: ShangGu-2 mirrors the multiple-sequence structure of the metal-poor disc.

The age information is sparse. Among ShangGu-2 members, only two stars have age estimates: one at 3.6 Gyr and one at 11.2 Gyr. The study notes explicitly that the 3.6 Gyr star has a $[\mathrm{Fe/H}] < -1.0$6 uncertainty larger than 0.2, so it does not appear in Fig. 16, while the 11.2 Gyr star is shown there as a black circle. Because the age sample is so small, the paper does not draw strong age conclusions from ShangGu-2 alone. This suggests that the chemical evidence is presently more informative than the age evidence for assessing the cluster’s origin (Hou et al., 21 Jul 2025).

5. Interpretation as a heated metal-poor disc component

The preferred interpretation of ShangGu-2 follows from the conjunction of its kinematics and chemistry. First, its $[\mathrm{Fe/H}] < -1.0$7 is similar to the cold disc rather than to a halo-like or clearly accreted population. Second, its $[\mathrm{Fe/H}] < -1.0$8 is unusually large, indicating substantial dynamical heating. Third, the presence of two $[\mathrm{Fe/H}] < -1.0$9–$> 1.4$0 branches, analogous to those in the metal-poor disc sample, argues against a single chemically distinct dwarf-galaxy progenitor. Fourth, its metallicity range remains within the broader regime in which the study argues for a chemically and dynamically complex metal-poor disc. Fifth, unlike ShangGu-1 and especially ShangGu-3, ShangGu-2 lacks a standout signature that would point to a separate progenitor.

For these reasons, the authors interpret ShangGu-2 as metal-poor disc stars that were dynamically heated, plausibly by merger-related or other perturbative processes. That interpretation is stated as the favored explanation rather than as a definitive proof. A plausible implication is that ShangGu-2 traces perturbative reshaping of an in-situ metal-poor disc population rather than the direct deposition of stars from a disrupted dwarf galaxy (Hou et al., 21 Jul 2025).

6. Relation to the other ShangGu substructures and diagnostic figures

The contrast with ShangGu-1 and ShangGu-3 clarifies what is distinctive about ShangGu-2. ShangGu-1 is characterized by $> 1.4$1, $> 1.4$2, metallicity generally below $> 1.4$3, and a positive correlation between orbital eccentricity and $> 1.4$4; it is interpreted as likely the remnant of an undiscovered dwarf galaxy. ShangGu-3 has disc-like rotation but very large radial motion, $> 1.4$5, mean $> 1.4$6, an $> 1.4$7-knee at $> 1.4$8, and a mean age of five stars of 13.1 Gyr with $> 1.4$9 Gyr; it is interpreted as likely coming from the GSE progenitor. Against this backdrop, ShangGu-2 is the most disc-like in rotation and the least suggestive of a separate accreted progenitor.

Substructure	Key signature	Interpretation
ShangGu-1	$(E, J_\phi, J_r, J_z)$0; $(E, J_\phi, J_r, J_z)$1; positive eccentricity–$(E, J_\phi, J_r, J_z)$2 correlation	Likely remnant of an undiscovered dwarf galaxy
ShangGu-2	$(E, J_\phi, J_r, J_z)$3; $(E, J_\phi, J_r, J_z)$4; two $(E, J_\phi, J_r, J_z)$5 sequences	Heated metal-poor disc stars
ShangGu-3	Disc-like rotation; $(E, J_\phi, J_r, J_z)$6; $(E, J_\phi, J_r, J_z)$7-knee at $(E, J_\phi, J_r, J_z)$8	Likely from the GSE progenitor

Several figures support the identification of ShangGu-2. Fig. 4 shows the DTGs in action/energy space and the grouping of unclassified DTGs into residual clusters. Fig. 14 compares the $(E, J_\phi, J_r, J_z)$9-$V_\phi \lesssim -200\ \mathrm{km\,s^{-1}}$0 and $V_\phi \lesssim -200\ \mathrm{km\,s^{-1}}$1-$V_\phi \lesssim -200\ \mathrm{km\,s^{-1}}$2 distributions for Clusters 4–6 and isolates ShangGu-2 as the cluster with cold-disc-like $V_\phi \lesssim -200\ \mathrm{km\,s^{-1}}$3 but large $V_\phi \lesssim -200\ \mathrm{km\,s^{-1}}$4. Fig. 16 shows $V_\phi \lesssim -200\ \mathrm{km\,s^{-1}}$5–$V_\phi \lesssim -200\ \mathrm{km\,s^{-1}}$6 for Cluster 5, ShangGu-2, against the kinematically cold disc and includes the auxiliary lines used earlier in the paper to highlight the two disc sequences (Hou et al., 21 Jul 2025).

7. Uncertainties, classification limits, and alternative readings

The study treats ShangGu-2 cautiously. The most immediate limitation is the sparseness of the age data: only two members have age estimates, which precludes a strong formation-time inference. The heated-disc interpretation is therefore supported mainly by kinematics and comparative chemistry rather than by a well-populated age distribution. The paper also does not claim chemical homogeneity for ShangGu-2; on the contrary, the existence of two $V_\phi \lesssim -200\ \mathrm{km\,s^{-1}}$7 sequences indicates internal complexity.

Methodological caveats are also explicit. HDBSCAN identifies dense regions and discards sparse regions as noise, so membership depends on the adopted clustering settings. The paper further stresses that kinematic boundaries between disc and non-disc populations are imperfect, so some contamination is expected. On that basis, the study does not rule out the possibility that some ShangGu-2 members are misclassified transition objects near the boundary between the cold and hot metal-poor disc. Accordingly, ShangGu-2 is best regarded as a dynamically selected substructure whose current interpretation is that of a heated disc population, but not yet a uniquely constrained formation channel (Hou et al., 21 Jul 2025).