Uncover 3D Dark Matter Distribution of the Milky Way by an Empirical Triaxial Orbit-Superposition Model: Method Validation

Basic Information

• Paper ID: 2506.23412
• Title: Uncover 3D Dark Matter Distribution of the Milky Way by an Empirical Triaxial Orbit-Superposition Model: Method Validation
• Authors: Ling Zhu, Xiang-Xiang Xue, Shude Mao, Chengqun Yang, Lan Zhang
• Classification: astro-ph.GA (Galactic Astrophysics)
• Publication Date: August 27, 2025
• Journal: Astronomy & Astrophysics
• Paper Link: https://arxiv.org/abs/2506.23412

Abstract

This paper proposes a novel dynamical model—an empirical triaxial orbit-superposition model—to investigate the three-dimensional distribution of dark matter in the Milky Way halo. The model is based on minimal physical assumptions, namely that the system is in a steady state, meaning that the six-dimensional phase-space distribution function of stellar orbits in the correct gravitational potential does not change with time. The authors validate the method by applying it to simulated datasets mimicking LAMOST + Gaia observations, successfully recovering the three-dimensional dark matter density distributions of three simulated galaxies with different intrinsic dark matter halo shapes.

Research Background and Motivation

Core Problem

The three-dimensional distribution of dark matter in the Milky Way, particularly its shape and density profile, is crucial for understanding the galaxy's formation history, yet significant uncertainties remain. Existing research faces the following challenges:

1. Inconsistent Observational Constraints: Different stellar stream studies provide contradictory constraints on dark matter halo shapes, with flattening parameter q values ranging from 0.44 to 1.3
2. Method Limitations:
   • Spherically symmetric Jeans models cannot constrain shape
   • Axisymmetric Jeans models rely on overly strict assumptions, and velocity ellipsoid orientations do not match observations
   • Existing distribution function models are mostly based on analytical assumptions, limiting flexibility

Research Significance

Accurately determining the three-dimensional distribution of the Milky Way's dark matter is important for:

• Understanding galactic formation and evolution history
• Validating cosmological simulation predictions
• Providing theoretical foundations for dark matter direct detection experiments

Technical Advantages

LAMOST + Gaia provides complete six-dimensional phase-space information (positions + velocities) for large samples of halo stars, offering unprecedented opportunities for constructing data-driven dynamical models.

Core Contributions

1. Proposed an empirical triaxial orbit-superposition model: First extension of the orbit-superposition method to triaxial modeling of the Milky Way halo, directly utilizing observed stellar orbits
2. Minimal physical assumptions: Only assumes steady-state systems, avoiding analytical distribution function assumptions in traditional methods
3. Data-driven weight determination: Orbital weights derived directly from selection function corrections rather than theoretical fitting
4. Simultaneous constraint of shape and density profile: Capable of recovering both the three-dimensional shape and radial density distribution of the dark matter halo
5. Method validation: Successfully validated the method's effectiveness on three Auriga simulated galaxies with different dark matter halo shapes

Methodology Details

Task Definition

Input: Six-dimensional phase-space observational data of halo stars (positions x, y, z and velocities vx, vy, vz) and their uncertainties Output: Three-dimensional density distribution ρ(x,y,z) of the dark matter halo and shape parameters (pDM, qDM) Constraint: System is in dynamical equilibrium

Model Architecture

1. Gravitational Potential Construction

Employs a three-component model:

• Bulge: Sérsic profile, Σbulge = Σ0 exp(-bn(Rgc/abulge)^(1/n))
• Disk: Miyamoto-Nagai model
• Dark Matter Halo: Flexible triaxial generalized NFW model

Dark matter density distribution:

ρhalo = ρ0(r̃/rs)^(-γ)[1 + (r̃/rs)^α]^((γ-β)/α) × exp(-(r̃/rcut)^ξ)

where r̃ = (pDMqDM)^(1/3)√(X² + (Y/pDM)² + (Z/qDM)²)

2. Orbit Integration

• Uses the AGAMA package for orbit integration
• Integrates 10 orbital periods for each star
• Samples 1000 particles at equal time intervals from each orbit
• Each particle inherits the weight of the initial star

3. Model Evaluation

Evaluates models by comparing distribution functions of model and observational data:

Density Distribution: χ²den = Σ(ρdata - ρmodel)²/dρ²data

Velocity Distribution: Computes likelihood of three velocity components in r-θ bins

• Lk = ΣNj i=1 log(Pki,j)
• χ²vdis = -2(Lvr + Lvϕ + Lvθ)

Total Evaluation: χ²tot = χ²den + χ²vdis

Technical Innovations

1. Orbit Library Construction: Directly uses observed stellar orbits rather than theoretical sampling
2. Weight Determination: Derived from selection function corrections rather than fitting optimization
3. Triaxial Modeling: Breaks through traditional axisymmetric limitations
4. Radial Variation: Allows shape parameters to vary with radius

Experimental Setup

Datasets

Auriga Simulations: Three Milky Way-mass cosmological hydrodynamic simulation galaxies selected

• Auriga 23: Primarily oblate dark matter halo (qDM ≈ 0.64, pDM ≈ 1.0)
• Auriga 5: Mildly triaxial shape (qDM ≈ 0.85, pDM ≈ 0.95)
• Auriga 12: Shape varies with radius, from oblate interior to prolate exterior

Simulated Observations:

• Distance error: 15%
• Tangential velocity error: 30 km/s
• Line-of-sight velocity error: 20 km/s
• Spatial selection: Simulates LAMOST northern hemisphere coverage
• Final sample: Approximately 20,000 halo stars

Data Processing

1. Substructure Removal: Uses position-velocity clustering to identify and remove substructure
2. Disk Star Removal: Based on joint criteria of orbital circularity λz and metallicity
3. Selection Function Correction: Density correction in the Rgc-zgc plane
4. Outlier Cleaning: 3σ clipping of velocity distribution

Evaluation Metrics

• Density Distribution Fitting: χ²den statistics
• Velocity Distribution Fitting: Log-likelihood of three velocity components
• Parameter Recovery Accuracy: Relative error compared to true values
• Shape Recovery: Precision of axis ratio constraints pDM, qDM

Experimental Results

Main Results

1. Radial Density Profile Recovery

• Auriga 23, 5, 12: Radial density profile recovery accuracy ~10% in the 4-50 kpc range
• Enclosed Mass: Relative uncertainty ~10%, consistent with true DM + gas mass
• Parameter Degeneracy: Significant degeneracy among ρ0, rs, γ, but radial profile well-constrained

2. Three-Dimensional Shape Constraints

Constant Shape Model:

• Auriga 23: qDM = 0.64±0.04, pDM = 1.00±0.04 (perfect agreement with true values qDM=0.64, pDM=1.0)
• Auriga 5: qDM = 0.85±0.06, pDM = 0.95±0.05 (essentially consistent with true values)
• Auriga 12: Captures outer region shape, but poor fit in inner region

Variable Shape Model (Auriga 12):

• Inner region: pin,DM > qin,DM ≈ 0.6 (oblate)
• Outer region: pout,DM < qout,DM ≈ 1.3 (prolate orientation)
• Successfully recovered the trend of shape variation with radius

Ablation Studies

1. Constraint Source Analysis

• Density Distribution: Primarily constrains dark matter shape
• Velocity Distribution: Primarily constrains radial density profile
• Radial Velocity vr: Most sensitive to mass distribution
• Stellar Halo Flattening qstar(r): Strongly constrains underlying dark matter shape

2. Orientation Constraints

• Tilt Angle βq: Fixed at 0° (perpendicular to disk)
• Azimuthal Angle αq: Weakly constrained due to data incompleteness
• Shape Degeneracy: Orientation degeneracy exists when pDM < qDM

Case Studies

Typical Success Case (Auriga 23)

• Best-fit model perfectly reproduces observed density distribution
• All three velocity components well-fitted
• Tight parameter constraints within 1σ confidence intervals

Challenging Case (Auriga 12)

• Constant shape model cannot simultaneously fit inner and outer regions
• Variable shape model significantly improves global fit
• Demonstrates method's adaptability to complex shape variations

Related Work

Limitations of Traditional Methods

1. Jeans Models: Spherical or axisymmetric assumptions limit shape constraint capability
2. Distribution Function Methods: Limited flexibility due to reliance on analytical DF assumptions
3. Stellar Stream Analysis: Inconsistent results, affected by LMC perturbations

Development of Orbit Modeling

1. Schwarzschild Method: Applied to nearby galaxies, requires theoretical orbit library
2. Made-to-Measure: Particle weights determined through fitting
3. Deep Learning Methods: Requires large data points, not yet applied to Milky Way

Innovations in This Work

• First extension of orbit superposition to Milky Way triaxial modeling
• Data-driven orbital weight determination
• Dynamical modeling with minimal physical assumptions

Conclusions and Discussion

Main Conclusions

1. Method Validity: The empirical triaxial orbit-superposition model successfully recovers the three-dimensional distributions of three different dark matter halos
2. Constraint Capability: Capable of simultaneously and accurately constraining both dark matter shape and radial density distribution
3. Strong Adaptability: Equally applicable to complex cases where shape varies with radius
4. Data-Driven: Avoids analytical assumptions in traditional methods, more objective

Limitations

1. Data Completeness: Requires high-quality six-dimensional phase-space data, currently applicable only to the Milky Way
2. Selection Function: Simplified selection function correction may introduce systematic errors
3. Orientation Degeneracy: Weak azimuthal angle constraints, degeneracy exists between pDM and αq
4. Equilibrium Assumption: Residual substructure and non-equilibrium features may affect results

Future Directions

1. Application to Real Data: Apply method to actual LAMOST + Gaia observations
2. Improved Selection Function: Adopt more precise three-dimensional selection function corrections
3. LMC Effects: Consider dynamical perturbations from the Large Magellanic Cloud
4. Gas Component: Include intergalactic medium in gravitational potential model

In-Depth Evaluation

Strengths

1. Method Innovation:
   • First proposal of data-driven triaxial orbit-superposition model
   • Breaks through traditional axisymmetric modeling limitations
   • Minimizes physical assumptions
2. Technical Advancement:
   • Fully exploits six-dimensional phase-space data from modern surveys
   • Flexible triaxial NFW model allows shape variation with radius
   • Rigorous statistical evaluation methodology
3. Sufficient Validation:
   • Validated on three different types of simulated galaxies
   • Includes complex cases with shape variations
   • Detailed ablation study analysis
4. Practical Value:
   • Provides new tools for Milky Way dark matter research
   • Directly applicable to LAMOST + Gaia data
   • Important implications for dark matter direct detection

Limitations

1. Computational Complexity:
   • Requires long-duration orbit integration for each star
   • Large computational cost for parameter space search
   • Scalability remains to be verified
2. Systematic Uncertainties:
   • Selection function correction method is relatively simplified
   • Insufficient assessment of residual substructure effects
   • Fixed disk and bulge parameters may introduce bias
3. Observational Constraints:
   • Only applicable to northern hemisphere data
   • Lack of data in inner regions affects constraints
   • Limited capability for orientation parameter constraints

Impact

1. Academic Contribution:
   • Opens new directions for Milky Way dynamical modeling
   • Provides new tools for validating cosmological simulations
   • Advances data-driven astrophysics
2. Practical Applications:
   • Directly applicable to current and future survey data analysis
   • Provides theoretical support for dark matter detection experiments
   • Helps understand Milky Way formation history
3. Method Generalization:
   • Extensible to modeling other galactic components
   • Provides methodological foundation for high-resolution observations of nearby galaxies
   • Promotes integration of observational astronomy and theoretical astrophysics

Applicable Scenarios

1. Milky Way Dark Matter Research: Using LAMOST, Gaia, and other survey data
2. Galaxy Formation Simulation Validation: Comparison with Illustris, EAGLE, and other simulations
3. Dark Matter Detection: Provides local density constraints for underground experiments
4. Nearby Galaxy Modeling: Analysis tool for future high-resolution observations

References

This paper cites important literature in the field, including:

• Auriga simulation series (Grand et al. 2017, 2019)
• Traditional Jeans modeling methods (Xue et al. 2008; Wang et al. 2020)
• Stellar stream constraint studies (Bovy et al. 2016; Vasiliev et al. 2021)
• LAMOST + Gaia data processing (Liu et al. 2017; Yang et al. 2022)

Summary: This is an important and innovative paper in the field of Milky Way dark matter research, proposing an empirical triaxial orbit-superposition model that provides a powerful new tool for constraining the three-dimensional distribution of dark matter using modern survey data. The method validation is thorough, the technical approach is clear, and it provides a solid foundation for subsequent practical applications.