Search for a Globular Cluster whose Passage through the Galactic Disk could Trigger the Radcliffe Wave

Based on a catalog of 152 globular clusters (GCs), their orbits were constructed to determine their intersections with the Galaxy’s plane of symmetry. Young open star clusters (OSCs) were selected from the selection zone characteristic of the Radcliffe wave. The Hunt and Reffert catalog served as the source of data on these OSCs. A grouping of 17 OSCs with an average age of 32.7 million years was found. It is compact in coordinate, velocity, and age space. This grouping is shown to be a good candidate for the hypothesis that the Radcliffe wave is generated by the passage of an impactor through the Galaxy’s plane of symmetry, with the impactor being the GC NGC 4372. The last time it crossed the galactic plane was 55.5 million years ago, and 22.2 million years later, a burst of star formation occurred at this location, forming a whole group of open-clustered stars, and possibly the Radcliffe wave as a whole. As a by-product, two relatively aged OSCs, HSC 1692 (58.0 million years) and HSC 1827 (68.0 million years), were found, the formation of which could have occurred after the intersection of the plane of symmetry of the Galaxy by the globular cluster Pal 13 85 million years ago.

Improvements to AI Systems Based on This Paper:

1. Temporal-Spatial Correlation Engine for Astrometric Data

• Improvement: Enhance AI models to jointly analyze 6D phase-space coordinates (position + velocity) and age distributions, not just clustering in 3D space.

• Capability: The improved system can automatically detect causally linked stellar groups (e.g., star formation bursts triggered by a passing object) by searching for compact clusters in coordinate-velocity-age hyperspace, even when the trigger (e.g., a globular cluster crossing) occurred millions of years earlier.

2. Predictive Impactor-Tracing Module

• Improvement: Add a module that back-propagates orbital trajectories of massive objects (like GCs) and cross-correlates their galactic plane crossings with local star formation timestamps.

• Capability: The AI can predict where and when a future or past impactor will trigger star formation, and identify candidate young clusters that should exist but have not yet been cataloged, enabling targeted observational follow-up.

3. Age-Uncertainty-Aware Clustering Algorithm

• Improvement: Incorporate age errors (e.g., ±5 Myr) into clustering loss functions, rather than treating ages as fixed values.

• Capability: The system can distinguish true co-eval stellar groups from chance alignments, reducing false positives in identifying star formation episodes triggered by external perturbations.

4. Multi-Source Catalog Fusion with Anomaly Detection

• Improvement: Train a transformer-based model to fuse heterogeneous catalogs (e.g., Hunt & Reffert for OSCs, plus GC orbit databases) and flag outliers like HSC 1692 and HSC 1827 as potential secondary impact remnants.

• Capability: The AI can automatically propose new hypotheses (e.g., a second impactor, Pal 13) by finding orphan clusters whose ages and positions match a different crossing event, without manual inspection.

5. Simulation-to-Observation Transfer Learning

• Improvement: Pre-train the AI on synthetic N-body simulations of galactic disk impacts, then fine-tune on real catalog data.

• Capability: The system can infer the mass, velocity, and crossing angle of an unseen impactor from the spatial and age distribution of resulting star clusters, providing physical parameters directly from observations.

6. Uncertainty-Quantified Causal Chain Generator

• Improvement: Implement a Bayesian network that links a GC’s plane crossing time to subsequent star formation bursts, outputting a probability for each causal link.

• Capability: The improved AI can generate a ranked list of plausible trigger–response pairs across the entire galaxy, quantifying confidence (e.g., 87% that NGC 4372 caused the Radcliffe wave) and suggesting which new observations would most reduce uncertainty.