🔭 astrophysics

Joint Modeling of GD-1 and C-19 as Old Streams

This paper uses DESI observational data to jointly model the GD-1 and C-19 stellar streams within a cosmological Milky Way-like potential, finding that their velocity spreads and density power spectra are best matched by Cold Dark Matter (CDM) subhalo simulations with progenitor cluster masses of approximately $8\times 10^4 M_\odot$ and $4\times 10^4 M_\odot$ , respectively, while ruling out Warm Dark Matter models and very young stream ages.

Original authors: Raymond G. Carlberg, Ting S. Li, Emma Jarvis, Nasser Mohammed, Joan Najita, Arjun Dey, Sergey E. Koposov, Jakob Piafsky, Leandro Beraldo e Silva, Constance M. Rockosi, J. Aguilar, S. Ahlen, D. Bianchi

Published 2026-06-16

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Raymond G. Carlberg, Ting S. Li, Emma Jarvis, Nasser Mohammed, Joan Najita, Arjun Dey, Sergey E. Koposov, Jakob Piafsky, Leandro Beraldo e Silva, Constance M. Rockosi, J. Aguilar, S. Ahlen, D. Bianchi, D. Brooks, T. Claybaugh, A. de la Macorra, Biprateep Dey, P. Doel, A. Font-Ribera, J. E. Forero-Romero, Satya Gontcho A Gontcho, G. Gutierrez, R. Joyce, S. Junea, A. Kremin, O. Lahav, M. Landriau, L. Le Guillou, A. Meisner, R. Miquel, W. J. Percival, C. Poppett, F. Prada, I. Pérez-Ràfols, G. Rossi, E. Sanchez, D. Schlegel, J. Silber, G. Tarlé, B. A. Weaver, R. Zhou, H. Zou

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: Cosmic Ghost Trails

Imagine the Milky Way galaxy as a giant, swirling city. Scattered throughout this city are "ghost trails" made of stars. These trails are the remains of ancient star clusters (groups of stars that were born together) that have been slowly torn apart by the galaxy's gravity over billions of years.

The paper focuses on two specific trails: GD-1 and C-19.

GD-1 is like a thin, tight ribbon of stars orbiting near the center of the galaxy.
C-19 is a more distant, looser trail orbiting near the very edge of the galaxy.

The astronomers wanted to figure out what kind of invisible "stuff" is floating around the galaxy that is messing with these trails. They suspected it was Dark Matter, but they wanted to know if it behaves like "Cold Dark Matter" (CDM) or "Warm Dark Matter" (WDM).

The Mystery: Why are the trails "hot"?

When a star cluster breaks apart, the stars should drift away slowly, like a group of friends walking out of a movie theater. They should stay relatively close together.

However, when astronomers looked at GD-1 and C-19, they found the stars were moving much faster and were more spread out than expected. The trails were "hot" (energetic) and "lumpy."

The Analogy: Imagine a line of people walking down a hallway.

The Smooth Hallway Theory: If the hallway is empty and smooth, the people stay in a neat line.
The Obstacle Theory: If the hallway is filled with invisible, bouncing balls (dark matter subhalos) that bump into the people, the line gets jumbled, people speed up, and the line gets wider.

The paper asks: Are the invisible balls "Cold" (heavy and clumpy) or "Warm" (lighter and fewer)?

The Experiment: A Time-Travel Simulation

To solve this, the researchers built a massive computer simulation. Think of it as a "Galactic Video Game" with a time machine.

Setting the Scene: They created a virtual Milky Way that evolves over 13 billion years. They included the main galaxy and a swarm of invisible dark matter "bumps" (subhalos) orbiting inside it.
The Start: They placed virtual star clusters in the simulation about 1 billion years after the Big Bang (when the universe was young).
The Run: They let the simulation run forward for 13 billion years, letting gravity tear the clusters apart and letting the invisible dark matter bumps bump into the stars.
The Comparison: They compared the resulting virtual trails to the real data collected by the DESI telescope (a giant camera that takes pictures of millions of stars).

The Key Findings

1. The "Cold" Theory Wins
The simulation showed that if the invisible dark matter is "Cold" (meaning there are lots of small, clumpy dark matter bumps), the virtual trails look exactly like the real GD-1 and C-19 trails. The stars get the right amount of "speed" and "spread."

If they used the "Warm" theory (where there are fewer, lighter bumps), the virtual trails stayed too cold and too neat. They didn't match the messy, fast-moving real trails.

The Verdict: The data strongly supports the idea that the galaxy is filled with many small, clumpy dark matter particles (Cold Dark Matter).

2. The Age of the Stars Matters
The researchers found that the trails had to be very old (about 12–13 billion years) to get the right amount of "heat." If they simulated younger trails (less than 10 billion years), the stars didn't have enough time to get bumped around enough to match reality.

3. How Big Were the Original Star Clusters?
By looking at how many stars are left in the trails, they estimated the size of the original "parent" star clusters before they were torn apart:

The parent of GD-1 was likely a massive cluster with about 80,000 suns.
The parent of C-19 was slightly smaller, with about 40,000 suns.

4. The "Power Spectrum" (The Lumpy Pattern)
The researchers didn't just look at how fast the stars were moving; they also looked at how "lumpy" the density of the stars was along the trail.

The Analogy: Imagine running your hand along a rope. Is it smooth, or does it have bumps and dips?
The "lumpiness" in the GD-1 trail matched the "Cold Dark Matter" simulation perfectly. The "Warm Dark Matter" simulation was too smooth. The real trail has 1.7 to 2.2 times more "bumps" than the Warm model predicts.

Summary

This paper is like a detective story where the clues are star trails. The detectives (astronomers) used a time-traveling computer simulation to see how invisible dark matter bumps would affect these trails over billions of years.

They concluded that the invisible dark matter in our galaxy is "Cold"—meaning it exists in many small, clumpy chunks that constantly bump into star trails, heating them up and making them wobble. This explains why the GD-1 and C-19 trails look the way they do today. The "Warm" theory, which suggests fewer and lighter dark matter chunks, doesn't fit the evidence.

Technical Summary: Joint Modeling of GD-1 and C-19 as Old Streams

Problem Statement
The paper addresses the challenge of understanding the kinematic structure of stellar streams, specifically GD-1 and C-19, within the context of the Milky Way's dark matter halo. While smooth, time-varying galactic potentials cannot explain the observed velocity dispersions of these streams (6–9 km s⁻¹), the interaction with dark matter subhalos is a proposed mechanism for "stream heating." However, previous models often lacked a complete dynamical history, creating degeneracies between the number of subhalos and the time streams have been orbiting. The authors aim to jointly model GD-1 and C-19 as old streams (formed ~1 Gyr after the Big Bang) evolving for 13 Gyr within a common, cosmologically motivated potential to determine if Cold Dark Matter (CDM) or Warm Dark Matter (WDM) subhalo populations better match observational data.

Methodology
The authors employ a simplified simulation framework designed to approximate detailed cosmological N-body simulations while allowing for the integration of star clusters over long timescales.

Potential and Subhalos: The simulation utilizes a time-dependent, multi-halo potential representing the assembly of a Milky Way-like galaxy. Three dominant halos merge over the first 7 Gyr to form a single dominant halo. Subhalos orbit within this potential, losing mass over time due to tidal stripping. The subhalo population is derived from cosmological simulations for both CDM and WDM (5.5 keV and 7 keV) models. The number density of subhalos declines over time, fitted to an exponential decay function, with distinct evolution rates for CDM and WDM.
Progenitor Clusters: Star clusters are modeled as Plummer spheres with half-mass radii of 5 pc (GD-1) and 3 pc (C-19). Clusters are inserted at $t \approx 1$ Gyr after the Big Bang. Stars are released from the cluster via a "particle spray" approach calibrated to N-body results, where stars become unbound based on the two-body relaxation rate and tidal forces.
Orbital Integration: To match observed orbits, the progenitor clusters are reverse-integrated from their current positions to the start time, accounting for the evolving potential and subhalo deflections. The simulations generate 36 quasi-independent realizations by rotating the initial subhalo distribution.
Data Comparison: The authors compare simulations to DESI observational data for GD-1 (620 probable members, reduced to 538 after cuts) and C-19 (41 members). Key metrics include:
- Velocity Width: The integrated line-of-sight velocity dispersion along the stream, modeled using Gaussian mixture models to separate the "core" and "cocoon" components.
- Longitudinal Density Power Spectrum: The Fourier power spectrum of density fluctuations along the stream length, which integrates over the stream width to avoid geometric biases.

Key Contributions and Results

Progenitor Mass Constraints: By matching the observed velocity widths and stream masses, the authors infer progenitor cluster masses of $\approx 8 \times 10^4 M_\odot$ for GD-1 and $\approx 4 \times 10^4 M_\odot$ for C-19. These masses are consistent with the requirement that clusters must be massive enough to survive the early chaotic phase of the galaxy to release stars at late times, forming the observed thin cores.
Stream Age and Heating: Simulations starting at 1 Gyr (resulting in 13 Gyr old streams) provide the best match to the data. Streams younger than 10 Gyr are ruled out as "insufficiently hot" (velocity widths too low). The velocity width of GD-1 increases significantly between 10 and 13 Gyr, whereas C-19 shows a more complex dependence due to its apocentric location.
CDM vs. WDM Discrimination:
- Velocity Width: CDM simulations provide the best match to the observed velocity widths for both streams. WDM simulations (5.5 keV and 7 keV) generally produce streams that are too cold (narrower velocity widths) because the subhalo mass function is truncated, reducing the number of low-mass subhalos capable of heating the stream.
- Density Power Spectrum: The longitudinal density power spectrum of the GD-1 stream is in good agreement with CDM simulations. The observed power is 1.7 to 1.9 times higher than that predicted by WDM 7 keV and 5.5 keV simulations, respectively. This indicates that the density fluctuations (gaps and clumps) driven by subhalo encounters are more pronounced in CDM.
Orbital Dependence: The study highlights that C-19, located near apocenter, exhibits a larger spread in simulation outcomes and is less sensitive to the specific dark matter model compared to GD-1, which is near pericenter.

Significance and Claims
The paper claims that joint modeling of GD-1 and C-19 within a single evolving cosmological potential successfully reproduces the observed kinematic properties of these old streams. The primary significance lies in the ability to distinguish between CDM and WDM models using stream heating metrics that are largely independent of geometric details (orbital blurring).

The authors assert that:

The observed velocity widths and density power spectra of GD-1 strongly favor a CDM subhalo population over WDM models with cutoffs at 5.5 or 7 keV.
The "cocoon" velocity width is a robust observable, independent of orbital blurring, providing a direct measure of the integrated subhalo heating history.
The progenitor masses derived ( $8 \times 10^4 M_\odot$ and $4 \times 10^4 M_\odot$ ) are necessary to explain the existence of the dense, cool cores observed today, implying that lower-mass clusters would have dissolved completely in the early universe.
While current data (particularly for C-19) limits the statistical confidence of the discrimination, the framework established allows for future improvements. The authors note that increasing the sample size of GD-1 stars with low velocity errors would significantly narrow the confidence range of the model matching.

The paper concludes that star streams serve as a sensitive probe of the starless dark matter subhalo population below the mass range of dwarf galaxies, and that the current DESI data combined with this modeling approach provides a viable path toward constraining the nature of dark matter.

The Big Picture: Cosmic Ghost Trails

The Mystery: Why are the trails "hot"?

The Experiment: A Time-Travel Simulation

The Key Findings

Summary

More like this