Lensing Tomography Without Rare Coincidences: Pseudo-Double Lenses vs Dark Energy
To measure cosmic expansion and dark energy with strong gravitational lensing, the double-source-plane lens technique is limited by rarity and mass-sheet degeneracy (MSD). This work presents pseudo double-source-plane lenses (PDSPLs) — pairs of independent single-source plane lenses whose deflectors have self-similar mass profiles. This approach generalizes the formalism to ~10^5 galaxy–galaxy lenses expected from LSST, Euclid, and Roman, eliminating the secondary MSD from intermediate source mass. Using a hierarchical forecasting framework that incorporates MSD, the study predicts constraints on the dark energy equation of state in a flat w0waCDM model. With LSST's 10-year photometric sample alone, σ(w0)≈0.45, while simultaneously determining MSD and power-law slope to ~2%. Adding a prior on Ωm (e.g., from CMB, BAO, SNe Ia) tightens σ(w0) to ~0.29, competitive with Stage III weak lensing. Notably, the large photometric volume outperforms smaller precisely measured spectroscopic samples, emphasizing statistical power over per-object precision.
Context
The nature of dark energy remains one of cosmology's greatest mysteries. Lensing tomography—comparing light deflection from sources at different distances—offers a geometric way to measure the expansion history of the universe, discovered by Edwin Hubble. However, in the strong lensing regime this method requires rare double lenses (two sources behind one deflector), which are extremely scarce. As early as 1937, Fritz Zwicky proposed using gravitational lenses to study distant objects; today his idea gets a new lease on life. With next-generation surveys like LSST, about 100,000 ordinary galaxy–galaxy lenses are expected, paving the way for a statistical approach and overcoming the rarity limitation.
Methods
The researchers simulated a population of gravitational lenses using the SLSim simulator, assuming deflectors as massive elliptical galaxies with a power-law density profile. To find nearly identical deflectors, a kD-tree algorithm was applied in the space of observed parameters: photometric redshift, effective radius, fluxes, and colors; for the spectroscopic subsample, precise redshifts and velocity dispersions (from spectroscopy data) were used. The resulting pairs were treated as pseudo-double lenses. In a Bayesian hierarchical analysis, cosmological parameters (including the dark energy equation of state) and deflector population characteristics—the mass-sheet degeneracy (MST) parameter and the density profile slope—were estimated simultaneously. Crucially, it was accounted that adding a constant mass sheet (modeling the contribution of dark matter) does not change image positions but distorts the mass scale—the famous degeneracy limiting the precision of lensing cosmography. A fixed Hubble constant reduced the dimensionality of the problem, as distance ratios are independent of the absolute scale.
Results
The simulations showed that the full 10-year LSST photometric sample (about 86,000 pairs) can measure the dark energy equation-of-state parameter w0 with a precision of σ(w0) ≈ 0.45 in a flat w0waCDM cosmology. Adding an external prior on the matter density Ωm (e.g., from supernova or cosmic microwave background data) improves the precision to σ(w0) ≈ 0.29, comparable to current weak lensing survey results. Meanwhile, the mean MST parameter and the deflector profile slope are determined to ~2% accuracy, and their intrinsic scatters with high confidence. Importantly, the massive photometric sample outperforms smaller spectroscopic subsamples, affirming the dominance of statistics over individual precision.
Implications
The pseudo-double lens method turns strong lensing from a hunt for unique configurations into a population-statistics tool. It requires no expensive velocity dispersion measurements and has systematic errors distinct from weak lensing, making it a valuable complement to other cosmological probes. The ability to simultaneously calibrate the mass-sheet degeneracy solves a longstanding problem that limited the precision of lensing cosmography.
Future development
Future machine-learning algorithms, such as contrastive learning, will be able to match deflector pairs based on their images, using hidden morphological features, improving matching accuracy. Accounting for galaxy evolution and external convergence will further refine the systematics. With the advent of Euclid and Roman surveys, the method can be extended to hundreds of thousands of lenses.
Impact
The results will influence the planning of LSST observation strategies and strong lensing data analysis, as well as the interpretation of future cosmological surveys.
Next steps
The next steps are to apply the method to real data, develop optimal metrics for pair matching that account for physical correlations, and conduct a full analysis of systematics, including selection effects.
Key open problems
This work is directly connected to the mass-sheet degeneracy problem in gravitational lensing and to determining the dark energy equation of state, one of the key unsolved challenges in modern physics.
🎯 The mass-sheet degeneracy is akin to adding a uniform invisible sheet of mass across the entire sky—it doesn't alter the lensing image but completely confounds the estimate of the deflector's true mass. Fritz Zwicky, the pioneer of gravitational lensing, hardly imagined such geometric elegance when he proposed using galaxies as lenses back in 1937.
Key numbers
- σ(w0) without Ωm prior: 0.45
- σ(w0) with Ωm prior: 0.29
- Number of LSST Y10 lenses: ~116 000
- Number of PDSPL pairs: ~86 000
- λMST determination precision: ~2%