---
title: Inverse Distance Ladder in Cosmology
url: https://www.emergentmind.com/topics/inverse-distance-ladder
type: topic
---

# Inverse Distance Ladder in Cosmology

Searching arXiv for recent and foundational papers on the inverse distance ladder.
Search query: "inverse distance ladder H0 BAO supernovae"
The inverse distance ladder is a family of cosmological distance-scale constructions that reverse the logic of the standard, locally anchored ladder. Instead of calibrating Type Ia supernovae from nearby geometric or stellar indicators and then extending outward, the inverse distance ladder starts from an absolute anchor at intermediate or high redshift—most commonly the BAO scale tied to the sound horizon, but in some implementations strong-lensing time-delay distances or late-time calibrators such as cosmic chronometers—and uses SN Ia as relative distance indicators to propagate that absolute scale toward \(z=0\). In current usage, the term therefore denotes not a single pipeline but a class of methods for inferring \(H_0\), reconstructing \(H(z)\), and testing the consistency of late- and early-Universe distance information [1411.1094][1905.12496][2406.05049][2505.22369][2602.12822].

## 1. Conceptual definition and reversal of the ladder

In the standard distance ladder, the calibration sequence is local and bottom-up: geometric or stellar calibrators set the absolute scale for secondary indicators, and those calibrated indicators are then pushed into the Hubble flow. Several papers in the inverse-ladder literature describe the method as the opposite construction: the scale is fixed at higher redshift and then carried downward through overlapping distance probes to infer the local expansion rate [1411.1094][2406.05049].

The canonical version combines BAO and SN Ia. In this arrangement, BAO act as standard rulers, SN Ia act as standard candles, and the overlap in redshift allows the two to calibrate one another. When the ruler length is supplied by the CMB-inferred sound horizon \(r_d\), BAO distances become absolute, SN Ia inherit that normalization, and \(H_0\) follows from extrapolation to \(z=0\) [1411.1094][1806.06781]. A distinct but conceptually parallel realization replaces the sound-horizon anchor with strong-lensing time-delay distances from lensed quasars, again using SN Ia to transport the absolute scale to low redshift [1905.12496].

A recurrent misconception is that the inverse distance ladder is necessarily a CMB-anchored BAO+SN method. The literature does not support such a restriction. Some analyses explicitly remove the usual \(r_d\) prior and use CC data to break the BAO–\(H_0\) degeneracy, while others construct a late-universe-only inverse ladder with SGL, CC, and GRB calibrators [2505.22369][2510.26355]. This suggests that “inverse distance ladder” is best understood as a methodological direction—calibration from higher to lower redshift—rather than as a unique dataset combination.

## 2. Core observables and mathematical structure

The SN component of the inverse ladder is fundamentally a relative-distance measurement. In one formulation, the theoretical SN distance modulus is written as
\[
\mu_{\rm theory}(z,\Theta)=5\log_{10}\!\left(\frac{D_L(z,\Theta)}{1\,{\rm Mpc}}\right)+25,
\]
while in SALT2-based analyses the observed relation is expressed as
\[
\mu = m_B - \left(M_B - \alpha X_1 + \beta C + \Delta_M\right).
\]
In both cases, the key degeneracy is the same: SN Ia strongly constrain the shape of the distance–redshift relation but not the absolute scale, because the absolute magnitude parameter remains unknown without an external anchor [2602.12822][1905.12496].

For BAO-based implementations, the central observables are the radial, transverse, and volume-averaged distances,
\[
D_H(z,\Theta)=\frac{c}{H(z,\Theta)}, \qquad D_M(z,\Theta)=\frac{D_L(z,\Theta)}{1+z},
\]
and
\[
D_V(z,\Theta)=\left(z\,D_M^2(z,\Theta)\,D_H(z,\Theta)\right)^{1/3}.
\]
BAO measurements constrain these quantities in units of the sound horizon, such as \(D_M/r_d\), \(D_H/r_d\), or \(D_V/r_d\), so an absolute \(H_0\) inference requires either a prior on \(r_d\) or an additional late-time calibrator [1411.1094][2406.05049][2602.12822].

In strong-lensing realizations, the anchor is the time-delay distance,
\[
D_{\Delta t} = (1+z_{\rm l}) \frac{D_{\rm d} D_{\rm s}}{D_{\rm ds}},
\]
which is approximately proportional to \(H_0^{-1}\). Time delays, lens mass modeling, and line-of-sight corrections yield posteriors for \(D_{\Delta t}\), and the SN Hubble diagram then transmits that absolute scale to \(z=0\) with substantially reduced background-model dependence relative to lensing alone [1905.12496].

A compact expression for the BAO+SN inverse-ladder likelihood is
\[
\chi^2(M_0,r_d,\Theta)=\chi^2_{\rm BAO}(r_d,\Theta)+\chi^2_{\rm SN}(M_0,\Theta),
\]
which makes explicit the separation between the absolute calibration parameters and the late-time expansion parameters [2602.12822].

## 3. Anchors and major implementations

The inverse distance ladder has been realized through several non-equivalent anchoring strategies.

| Anchor type | Representative data combination | Characteristic role |
|---|---|---|
| CMB-inferred \(r_d\) | BAO + SN Ia | Standard-ruler calibration of the SN Hubble diagram |
| Time-delay distances | H0LiCOW/TDCOSMO lenses + SN Ia | Absolute distance anchor at intermediate redshift |
| Late-time calibrators without CMB \(r_d\) prior | BAO + SN Ia + CC, or BAO + SN Ia + SGL + CC + GRB | Breaks BAO–\(H_0\) degeneracy using late-time data |

The CMB-anchored BAO+SN construction is the most widely used form. In the traditional version summarized in several papers, a prior on \(r_d\) derived from the CMB or BBN anchors BAO, BAO calibrate SN Ia, and the resulting distance relation is extrapolated to \(z=0\) to infer \(H_0\) [1411.1094][1806.06781][2505.22369]. Its precision is high because the sound horizon is tightly constrained, but its absolute scale is inherited from early-Universe physics.

The strong-lensing version substitutes a geometrical distance anchor for the sound horizon. Using four H0LiCOW quasar lenses—B1608+656, RXJ1131−1231, HE0435−1223, and SDSS 1206+4332—one study showed that combining lensing with the JLA SN sample largely removes the cosmological-model sensitivity seen in lensing-only \(H_0\) inference [1905.12496]. This implementation is notable because it does not rely on Cepheids and does not anchor the distance scale with the early-Universe ruler.

A third class of analyses explicitly aims to avoid a CMB- or BBN-based \(r_d\) prior. One “improved inverse distance ladder” uses DESI DR2 BAO, CC data, and either DESY5 or PantheonPlus SN samples, with the PAge parameterization supplying a global late-time expansion history [2505.22369]. A related “PAge-improved IDL” uses DESI DR2 BAO and DESY5 SNe, calibrated by SGL, CC, and GRB, and is described as a late-universe-only construction [2510.26355]. These variants show that inverse-ladder logic does not require the sound horizon to enter as an external prior.

## 4. Reconstruction strategies and degrees of model dependence

A major theme of the literature is that inverse-ladder results depend not only on the anchor but also on how the low-redshift expansion history is represented. One widely used strategy is cosmography. An updated DES analysis fits third-, fourth-, and fifth-order cosmographic models to DES-SN5YR and DESI BAO, with
\[
H(z)=H_0\left[1+\mathcal H_1 z+\mathcal H_2 z^2+\mathcal H_3 z^3+\mathcal H_4 z^4\right],
\]
and a corresponding luminosity-distance series. With the inclusion of higher-redshift DESI BAO, the third-order model is reported to be a poor fit, while the fourth-order model is preferred by the Akaike Information Criterion [2406.05049].

A different route is flexible parametric reconstruction of \(H(z)\). One model-independent reconstruction uses BAO, Pantheon SN Ia, and \(r_d\) priors to fit “epsilon” and “log” parameterizations of \(H^2(z)\), emphasizing that the resulting \(H(z)\) constraints are independent of detailed dark-sector physics at low redshift and rely only on the validity of the FRW metric of GR [1806.06781]. Another set of analyses uses PAge and MAPAge, parameterizations based on cosmic age rather than a specific dark-energy model, with the stated motivation that \(H(t)t\) evolves more smoothly than \(H(z)\) [2505.22369][2510.26355].

A more radical attempt to reduce global-model dependence uses the distance-duality relation
\[
d_L=(1+z)^2 d_A,
\]
valid if photon number is conserved and gravity is described by a metric theory. In this construction, SN data are binned at BAO redshifts and compared directly to BAO-derived luminosity distances at the same redshift, so that no model is adopted to calibrate BAO with supernovae at \(z>0.15\) [1910.14125].

Inverse-ladder methods have also been used in nonstandard diagnostics. A model-independent anisotropy test reconstructs \(d_L(z)\) by GP, using Pantheon+ SN Ia for relative distances and H0LiCOW lensed quasars for absolute anchoring. In that work, the inverse ladder is not used to obtain a new \(H_0\) value, but to compare sky-region reconstructions of the luminosity-distance relation [2407.19278].

## 5. Empirical determinations of \(H_0\)

Representative inverse-ladder determinations of \(H_0\) span a broad range because the absolute anchor and the reconstruction method differ materially across analyses.

| Study | Implementation | Reported \(H_0\) |
|---|---|---|
| "Calibrating the cosmic distance scale ladder" [1411.1094] | BAO + SN + CMB sound-horizon anchor | \(67.7 \pm 1.1\) |
| "Model independent \(H(z)\) reconstruction using the cosmic inverse distance ladder" [1806.06781] | BAO + Pantheon + Planck \(r_d\) prior | \(68.42 \pm 0.88\) |
| "The Dark Energy Survey Supernova Program" [2406.05049] | DES-SN5YR + DESI BAO + fourth-order cosmography | \(67.19^{+0.66}_{-0.64}\) |
| "On The Stability Of \(H_0\) And The Inverse Distance Ladder" [2602.12822] | DES-Dovekie SN + DESI BAO + CMB \(r_d\) prior | \(66.82 \pm 0.74\) |
| "The Hubble Constant determined through an inverse distance ladder including quasar time delays and Type Ia supernovae" [1905.12496] | H0LiCOW lenses + JLA SN Ia | \(73\)–\(74\) |
| "Model-independent cosmological inference after the DESI DR2 data with improved inverse distance ladder" [2505.22369] | DESI DR2 BAO + CC + DESY5 SN in PAge | \(67.91 \pm 2.33\) |
| "Model-independent late-universe measurements of \(H_0\) and \(Ω_\mathrm{K}\)" [2510.26355] | DESI + DESY5 + SGL + CC + GRB in MAPAge | \(71.59 \pm 0.94\) |

The CMB-anchored BAO+SN branch of the literature is notably stable around \(H_0 \simeq 67\)–\(68\). The 2026 stability study states that current inverse-distance-ladder analyses cannot explain the Hubble tension when anchored to the CMB, and further argues that future inverse-distance-ladder measurements anchored to current CMB data will remain near \(H_0 \simeq 66.5 \pm 0.5\) [2602.12822]. This conclusion is reinforced by the updated DES analysis, which finds a low \(H_0\) without assuming flat \(\Lambda\)CDM, even though its best-fitting expansion history differs from Planck’s [2406.05049].

By contrast, the strong-lensing+SN implementation gives \(H_0\) consistently around \(73\)–\(74\) km s\(^{-1}\) Mpc\(^{-1}\) across flat and non-flat \(\Lambda\)CDM, \(w\)CDM, and \(w_0w_a\)CDM, with only about \(2\%\) variation across the tested cosmologies [1905.12496]. This is an important empirical divergence within the inverse-ladder family itself.

Late-universe-only variants occupy an intermediate position. The improved IDL with DESI DR2 BAO, CC, and DESY5 gives \(H_0 = 67.91 \pm 2.33\), consistent with Planck and in \(2.0\sigma\) tension with SH0ES [2505.22369]. The PAge-improved late-time IDL with DESI, DESY5, SGL, CC, and GRB yields \(H_0=71.59\pm0.94\) in MAPAge, reducing the SH0ES tension to the \(1.0\sigma\) level [2510.26355]. Taken together, these results indicate that inverse-ladder inferences are not monolithic; the anchor choice dominates the final absolute scale.

## 6. Controversies, diagnostics, and statistical systematics

One of the strongest claims in the recent literature is that known SN Ia systematics do not provide enough freedom to move a CMB-anchored inverse-ladder result from \(H_0 \sim 67\) to the local-ladder regime near \(73\)–\(74\). Re-fitting DES-Dovekie SN variants changes the inferred \(H_0\) by \(<0.1\) km s\(^{-1}\) Mpc\(^{-1}\), and the redshift-dependent magnitude evolution required to shift the inverse ladder to the local value is reported as \(d\mu/dz \simeq 0.2\) mag. In flat \(\Lambda\)CDM\), that change would drive \(\Omega_M\) to \(0.23\), \(10\sigma\) discrepant with other cosmological probes [2602.12822].

Another controversy concerns BAO self-consistency. A distance-duality-based inverse-ladder construction reported strong inconsistency between angular-only BAO constraints and anisotropic BAO measurements. In that analysis, SNe+\(\theta\) BAO plus a CMB prior on \(r_d\) gave \(H_0=74.36\pm1.42\), whereas SNe+\(\alpha_\perp\) BAO plus the same anchor gave \(H_0=69.71\pm1.28\). The authors concluded that clarifying the tension between angular and perpendicular anisotropic BAO is a necessary step toward understanding the \(H_0\) crisis [1910.14125].

Inverse-ladder methods have also been used to test isotropy. A GP reconstruction based on Pantheon+ and H0LiCOW found that \(d_L(z)\) reconstructions from different Galactic regions are almost consistent with each other, with only a very weak preference for cosmic anisotropy. The globally fitted SN absolute magnitude was reported as \(M=-19.2522^{+0.0270}_{-0.0279}\) at \(68\%\) CL [2407.19278].

The statistical foundations of ladder calibration matter as well. A 2025 study argued that distance-ladder inference is a prior-and-selection problem, emphasizing that a flat prior in distance modulus implies \(\pi(r)\propto 1/r\), whereas a homogeneous population suggests \(\pi(r)\propto r^2\). The paper explicitly extends the point to inverse distance ladders, arguing that any bias or mis-modeling in the ladder calibration stage propagates into the combined cosmological inference [2511.03394]. A related but more phenomenological meta-analysis of \(H_0\) determinations classified inverse distance ladder mainly as a hybrid or comparison method rather than a separate final category, and argued that the central divide in the literature is between distance-ladder measurements and non-distance-ladder measurements [2408.11031].

Finally, inverse-ladder constraints remain entangled with the quality of local-ladder calibration. A study of Cepheid outliers found that its outlier treatment increased the uncertainty in Supernova-host distances by a median factor of \(\sim 30\%\), added in quadrature \(1.2\) km s\(^{-1}\) Mpc\(^{-1}\) to the statistical uncertainty of \(H_0\), and led to \(H_0=72.6\pm2.8\) km s\(^{-1}\) Mpc\(^{-1}\), a value stated to be fully consistent with both the Planck and inverse-distance-ladder \(H_0\) constraints [1507.07523]. This suggests that the inverse ladder functions not only as an alternative route to \(H_0\), but also as a cross-check on how local-ladder systematics propagate into cosmological calibration.

Source: https://www.emergentmind.com/topics/inverse-distance-ladder