Dark Sector Memory Test (DSMT)

Probing Dark Matter and Dark Energy History-Dependence via Cluster Mergers

A model-agnostic observational framework for testing kinematic-history dependence in gravitational lensing

---

🔭 Overview

The Dark Sector Memory Test is a model-agnostic observational framework for testing whether gravitational lensing encodes kinematic history beyond the instantaneous matter distribution.

Multiple theoretical frameworks predict "memory" effects in the dark sector:

Theory	Mechanism	Prediction
Superfluid Dark Matter	Phase transitions at critical velocity	Turbulent wakes (Sivakumar+ 2025)
Non-Markovian EFT	Memory kernels from heavy fields	Smooth power-law response (Chaudhuri+ 2025)
Nonlocal Gravity	Retarded stress-energy coupling	Logarithmic response (Maggiore+ 2014)
ΛCDM	No history dependence	No correlation (null hypothesis)

The key question: Do lensing convergence residuals correlate with merger infall velocity?

Despite theoretical predictions, no systematic observational test has been performed — until now.

---

📊 Pilot Analysis Results (January 2026)

Current Sample

We have analyzed three merging galaxy clusters with Hubble Frontier Fields convergence maps and published kinematic constraints:

Cluster	z	v_infall (km/s)	Mach	Geometry	Kinematic Source
MACSJ0416	0.396	1600 ± 400	—	Plane-of-sky	Jauzac+ 2015
Abell 2744	0.308	2000 ± 200	1.2	Plane-of-sky	Chadayammuri+ 2024
Abell 370	0.375	~3000 (LOS)	—	Line-of-sight	Bimodal redshifts

Key Finding: Geometry Dependence

Plane-of-sky mergers show a positive correlation between infall velocity and residual dipole moment:

Cluster	v (km/s)	Dipole	Asymmetry	Residual RMS
MACSJ0416	1600	0.037	0.886	0.066
Abell 2744	2000	0.044	0.934	0.087

Abell 370 deviates from this trend — but this is physically expected: its merger is largely along the line of sight, so any wake signature would be foreshortened in projection.

Preliminary Interpretation

• For plane-of-sky mergers, higher infall velocity → larger dipole moment and asymmetry
• This is consistent with wake signature predictions from superfluid DM and nonlocal gravity
• Line-of-sight mergers require projection corrections before inclusion

Next Steps

• Abell 2146 (v = 2700 km/s, Mach = 2.3 ± 0.2) — best-constrained kinematics from shock measurements (Russell+ 2012). Convergence map requested from Coleman/King.
• With 3+ plane-of-sky mergers, we can compute statistically meaningful correlations

---

🎯 What This Project Does

1. Reconstructs lensing convergence maps from HST weak+strong lensing data
2. Constructs ΛCDM baseline models via parametric fitting (Gaussian smoothing)
3. Computes residual maps: Δκ = κ_obs − κ_baseline
4. Measures six morphological metrics quantifying residual structure
5. Tests for correlations between metrics and published merger kinematics

---

🚀 Quick Start

Installation

git clone https://github.com/LBCplus/Dark-Sector-Memory-Test.git
cd Dark-Sector-Memory-Test
pip install -r requirements.txt

Download Data

# Download Frontier Fields convergence maps from MAST
python code/download_and_analyze.py --download --data-dir ./data

Run Analysis

python code/download_and_analyze.py --analyze --data-dir ./data

---

📁 Repository Structure

Dark-Sector-Memory-Test/
├── README.md                       # You are here
├── LICENSE                         # MIT License
├── CITATION.cff                    # Citation metadata
├── requirements.txt                # Python dependencies
│
├── paper/
│   ├── dsmt_paper_draft.md         # Manuscript draft
│   └── figures/                    # Publication figures
│
├── code/
│   ├── dsmt_analysis.py            # Main analysis module (~700 lines)
│   └── download_and_analyze.py     # Data pipeline with published kinematics
│
├── docs/
│   ├── methodology.md              # Detailed methods
│   └── statistical_analysis_plan.md # Pre-specified analysis
│
├── configs/
│   └── pilot_study.yaml            # Analysis configuration with kinematic params
│
└── data/                           # Data directory (not tracked)
    └── .gitkeep

---

📐 Morphological Metrics

We quantify lensing residual structure using six metrics:

Metric	Symbol	Definition	Interpretation
Dipole moment	|d|	∫ Δκ(x) x d²x	Preferred direction of excess mass
Quadrupole	Q	Eigenvalue ratio of Q_ij	Elongation of residuals
Tail-alignment	T	cos(2(θ_res − θ_merger))	Alignment with merger axis
Asymmetry	A	Σ|Δκ − Δκ_180°| / 2Σ|Δκ|	Departure from point symmetry
Centroid offset	|Δx_c|	|x_obs − x_baseline|	Mass center displacement
Power spectrum	P_tot	∫ P(k) dk	Total residual structure

---

📊 Pilot Sample

Cluster	z	v_infall (km/s)	Data Source	Status
MACSJ0416	0.396	1600 ± 400	HFF CATS v4	✅ Analyzed
Abell 2744	0.308	2000 ± 200	HFF CATS v4.1	✅ Analyzed
Abell 370	0.375	~3000 (LOS)	HFF CATS v4	✅ Analyzed (LOS geometry)
Abell 2146	0.232	2700 (+400/−300)	Coleman+ 2017	⏳ Data requested

Kinematic Sources

• MACSJ0416: Jauzac et al. 2015, MNRAS 446, 4132
• Abell 2744: Chadayammuri et al. 2024, arXiv:2407.03142 (2.1 Ms Chandra + JWST)
• Abell 370: Bimodal galaxy redshift distribution (~3000 km/s separation)
• Abell 2146: Russell et al. 2012, MNRAS 423, 236 (shock Mach numbers)

---

🎓 Theoretical Background

Why "Memory"?

Standard ΛCDM predicts that lensing depends only on the current matter distribution. Several beyond-ΛCDM theories predict dependence on kinematic history:

Superfluid Dark Matter (Berezhiani & Khoury 2015)

"Merger dynamics depend on the infall velocity versus phonon sound speed; distinct mass peaks in bullet-like cluster mergers correspond to superfluid and normal components."

Sivakumar et al. (2025) — Most direct prediction:

"Merger-induced turbulence should produce asymmetric, fine-structure residuals in lensing maps, correlated with infall velocity."

Non-Markovian EFT (Chaudhuri et al. 2025)

Memory kernels from integrated-out heavy fields produce history-dependent gravitational response.

Nonlocal Gravity (Maggiore & Mancarella 2014)

Past stress-energy contributes to present gravitational dynamics via retarded Green's functions.

The Gap We Address

Cognola et al. (2022) tested nonlocal gravity against cluster lensing and found it indistinguishable from GR using standard methods, concluding that "a different discriminator is needed."

DSMT is that discriminator.

---

📚 Key References

Theoretical Foundations

• Berezhiani & Khoury (2015) — Superfluid DM framework — PRD 92, 103510
• Sivakumar et al. (2025) — Turbulent mergers — PRD 111, 083511
• Chaudhuri et al. (2025) — Non-Markovian EFT — arXiv:2509.22293
• Maggiore & Mancarella (2014) — Nonlocal gravity — PRD 90, 023005

Observational Data

• Grayson et al. (2024) — MACSJ0416 BUFFALO model — MNRAS 536, 2690
• Russell et al. (2012) — Abell 2146 kinematics — MNRAS 423, 236
• Chadayammuri et al. (2024) — Abell 2744 multiwavelength — arXiv:2407.03142
• Coleman et al. (2017) — Abell 2146 strong lensing — MNRAS 464, 2469

Gap Identification

• Cognola et al. (2022) — Nonlocal gravity vs. GR degeneracy — arXiv:2205.03216

---

🤝 Contributing

Contributions welcome! Particularly interested in:

• Additional cluster convergence maps with published kinematic constraints
• Simulation comparisons (TNG-Cluster, BAHAMAS)
• Statistical methodology improvements
• Theoretical predictions from other frameworks

Please open an issue or submit a pull request.

---

📄 Citation

If you use this code or methodology, please cite:

@software{dsmt2026,
  author = {[Author]},
  title = {Dark Sector Memory Test: Probing Dark Matter and Dark Energy History-Dependence via Cluster Mergers},
  year = {2026},
  url = {https://github.com/LBCplus/Dark-Sector-Memory-Test}
}

---

📜 License

This project is licensed under the MIT License — see LICENSE for details.

---

🙏 Acknowledgments

• HST Frontier Fields and BUFFALO teams for public lensing data
• MAST archive for data hosting
• Chandra X-ray Observatory for archival data
• The lensing community for published convergence maps

---

Testing whether gravity remembers