Mass-Sheet Transformation in Gravitational Lensing

Updated 11 January 2026

Mass-sheet transformation is a continuous symmetry in strong gravitational lensing that preserves image positions, shapes, and flux ratios while uniformly scaling magnifications and time delays.
It introduces a degeneracy in lens mass profile reconstructions that affects cosmological inferences, particularly the determination of the Hubble constant from time-delay measurements.
Breaking the MST requires external constraints like stellar velocity dispersion, weak lensing data, or magnification bias to achieve accurate lens modeling and robust cosmological parameters.

The mass-sheet transformation (MST) is a continuous symmetry of strong gravitational lensing that generates a family of lens mass distributions differing by a uniform mass sheet and an overall scaling. This transformation exactly preserves all image positions, shapes, and flux ratios in geometric optics, but rescales absolute magnifications, time delays, and hence any physical quantities dependent on the overall lensing potential normalization. The MST forms the core of a broader degeneracy, the source-position transformation, and is central to the interpretation of cosmological inferences from lensing, such as measurements of the Hubble constant.

1. Mathematical Formulation and Invariance Properties

In the standard thin-lens approximation, the observed position $\theta$ of a lensed image and the (unobservable) source position $\beta$ are related by

$\beta = \theta - \alpha(θ),$

where $\alpha(θ)$ is the deflection angle, itself derived from the lensing potential $\psi(\theta)$ ,

$\alpha(\theta) = \nabla\psi(\theta).$

The surface mass density is parametrized as the convergence

$\kappa(\theta) = \frac{\Sigma(\theta)}{\Sigma_{\mathrm{crit}}},$

with $\Sigma_{\mathrm{crit}}$ the usual critical density.

The MST is defined by a global one-parameter mapping: $\kappa(\theta) \rightarrow \kappa_\lambda(\theta) = \lambda\,\kappa(\theta) + (1-\lambda),$

$\psi(\theta) \rightarrow \psi_\lambda(\theta) = \lambda\,\psi(\theta) + \frac{1}{2}(1-\lambda)|\theta|^2,$

$\alpha(\theta) \rightarrow \alpha_\lambda(\theta) = \lambda\,\alpha(\theta) + (1-\lambda)\theta,$

where $\lambda\in\mathbb{R}$ parametrizes the transformation.

The transformed lens equation becomes

$\beta_\lambda = \theta - \alpha_\lambda(\theta) = \lambda[\theta-\alpha(\theta)] = \lambda\,\beta,$

which shows that the image angles $\theta$ solving the original equation correspond to the same observed image positions, with the only effect being a scaling of the source plane $\beta \to \lambda\beta$ (Schneider et al., 2013, Unruh et al., 2016).

Key invariants and scalings under MST:

Image positions, relative shapes, and flux ratios remain exactly unchanged.
Time delays: For arrival-time differences (Fermat potential), the difference rescales,

$\Delta t_{ij} \to \lambda\,\Delta t_{ij} + \text{const},$

thus any inferred time-delay distance is degenerate under $\lambda$ (Wertz et al., 2017, Chen et al., 2020).

Magnifications: $\mu\to\mu/\lambda^2$ .
Einstein radius and projected mass interior to it are conserved: $M_\lambda(<\theta_E) = M(<\theta_E)$ (Unruh et al., 2016).

2. MST as a Special Case of the Source-Position Transformation

The MST forms a one-parameter subgroup of the broader source-position transformation (SPT), which comprises all smooth, invertible mappings $s:\,\beta\to\hat{\beta}=s(\beta)$ . The general SPT induces a new deflection law,

$\hat\alpha(\theta) = \alpha(\theta) + \beta(\theta) - s(\beta(\theta)),$

so that the image positions remain invariant under $(\alpha,\beta)\to(\hat\alpha, s(\beta))$ , i.e., the same $\theta$ solve both equations (Wertz et al., 2018).

MST is recovered when $s_{\rm MST}(\beta)=\lambda\beta$ , yielding the standard MST formulas.

Table 1: Relationship Between SPT and MST (special case)

Transformation	SPT	MST (special case)
Source map	$s(\beta)$	$s_{\rm MST}(\beta)=\lambda\beta$
Deflection	General $\hat\alpha(\theta)$	$\hat\alpha(\theta) = \lambda\alpha(\theta) + (1-\lambda)\theta$

3. Applications, Degeneracies, and Impact on Cosmological Parameters

The invariance of image configurations under the MST introduces a degeneracy in the mass profile reconstruction—commonly referred to as the "mass-sheet degeneracy"—limiting the ability to uniquely determine absolute masses and cosmological distances from strong lens data alone (Khadka et al., 2024, Chen et al., 2020, Poon et al., 2024).

For time-delay cosmography:

Hubble constant $H_0$ : The observable time-delay distance $D_{\Delta t}$ is degenerate under MST:

$D_{\Delta t}^{(\lambda)} = D_{\Delta t} / \lambda,$

resulting in $H_0$ inferred from lensing being inversely proportional to $\lambda$ . Without external constraints, lensing-only measurements cannot achieve percent-level accuracy in $H_0$ (Chen et al., 2020, Poon et al., 2024, Teodori et al., 2022).

Gravitational wave (GW) lensing: The MST causes effective GW luminosity distances to scale by $\lambda$ with a full degeneracy between $d_L$ and the sheet parameter. Even when adding EM constraints such as lens redshift and Einstein radius, only the combination $H_0 \cdot \lambda$ can be inferred. Only with independent velocity-dispersion (stellar kinematic) measurements can $\lambda$ be directly constrained (Poon et al., 2024).

For multi-plane lensing systems, the degeneracy generalizes to a sequence of isotropic rescalings and sheet additions in each lens plane, preserving all lensing observables except for an overall scaling of time delays in all source planes (Schneider, 2014, Schneider, 2014).

4. Breaking the Mass-Sheet Degeneracy: Methods and Practical Strategies

To break the MST, it is necessary to include observables or external data that do not transform covariantly under the MST. Key methods include:

Stellar velocity dispersion: Kinematic data provide an independent mass normalization at the lens, allowing $\lambda$ to be measured, as the observable velocity dispersion $\sigma$ scales as $\sigma^2\propto\lambda$ (Chen et al., 2020, Poon et al., 2024).
Weak lensing (galaxy-galaxy lensing): Measurements of the tangential shear and differential surface mass density at large radii directly constrain $\lambda$ since $\gamma_t(R)\to\lambda\gamma_t(R)$ under MST, providing $1-10\%$ precision on the sheet parameter with large survey data (Khadka et al., 2024, Rexroth et al., 2016).
Number-count magnification (magnification bias): Magnification statistics from background number counts can constrain the overlying mass sheet when combined with shear/flexion data (Rexroth et al., 2016).
Lens environment and external convergence: Line-of-sight structures (external convergence) need to be estimated via cosmological simulations, galaxy counts, or weak lensing, and must be included with appropriate nuisance parameterization (Teodori et al., 2022).
Model-independent distance ratios: For time-delay cosmography, combining SN Ia and BAO data allows the required $D_s/D_{ds}$ ratio to be inferred without explicit cosmological priors, constraining $\lambda$ in a model-independent manner (Chen et al., 2020).

5. MST in Gravitational Wave Lensing and Interference Regimes

Mass-sheet degeneracy also affects strongly lensed GW observations in the geometric-optics regime, where only ratios of time delays and image positions are observable. However, in the interference regime (wave periods comparable to lensing time delays), the MST modifies interference patterns in the GW amplification factor, enabling direct measurement of $\lambda$ —a single lensed GW detection in this regime can break the MST, with simulation-based studies forecasting $1\sigma$ uncertainties in $\lambda$ of $\sim12\%$ with current detector sensitivity, and down to $2\%$ for future high-SNR events (Cremonese et al., 2021).

For galaxy lens reconstruction with GWs alone ("dark sirens"), the mass-sheet degeneracy renders GW luminosity distance and time-delay scale fully degenerate with the MST parameter. Complete lens modeling and $H_0$ measurement thus require EM follow-up, particularly velocity dispersion measurements (Poon et al., 2024).

6. Extensions: Multi-Source and Multi-Plane Systems, and Line-of-Sight Effects

In multi-plane and multi-source lens systems, a generalized MST exists, acting with plane-dependent scaling and additive tidal matrices. All relative image observables and time delays in each source plane rescale by a common factor, so cosmological parameter estimation is again fundamentally degenerate (Schneider, 2014, Schneider, 2014).

Multiple sources or images at different redshifts do not break the global MST, since only differential combinations of line-of-sight convergence are measurable (Teodori et al., 2022). Internal (core-like) convergence degeneracies complicate the interpretation further, amplifying the need for comprehensive nuisance parameter marginalization over both internal and external mass sheet components.

7. Physical Interpretation and Theoretical Insights

The MST can be interpreted optically as a scaling symmetry of the "image-selection relation" (ISR), which isolates the lensing geometrical focusing term and describes all image-forming content. The geometric focusing (uniform mass sheet) adds to the inhomogeneous image-selection lens (ISL). The scaling symmetry of the ISR underlies the MST: multiplying the ISL deflection and source position by $\varepsilon$ leaves all image positions unchanged but rescales magnifications by $\varepsilon^{-2}$ and time delays by $\varepsilon$ (Gorenstein, 4 Jan 2026). Restoring the geometric focusing yields exactly the MST, demonstrating the fundamental connection between optical invariance and the mass-sheet degeneracy.