Tests of Evolving Dark Energy with Geometric Probes of the Late-Time Universe
Kunhao Zhong, Bhuvnesh Jain · 2025
Read Tests of Evolving Dark Energy with Geometric Probes of the Late-Time Universe by Kunhao Zhong, Bhuvnesh Jain (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-26480.txt.
Introduction
Kunhao Zhong and Bhuvnesh Jain test whether dark energy really changes over time, in work from 2025. Recent survey results favor a dark energy that evolves, but the claim rests on combining several kinds of data.
The authors isolate the pure geometry in weak lensing data from the Dark Sky survey's third year, setting aside growth effects with an extra parameter. Combined with distance measures and early-universe data, the evolving model wins at about three standard deviations. Adding supernova data makes the answer fragile: leaving out the nearest supernovae, or using only distant ones, drops the support below two. Future lensing surveys should match today's supernova power.
Catalog record
| CREATOR | Kunhao Zhong, Bhuvnesh Jain |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 37 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-26480
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Tests of Evolving Dark Energy with Geometric Probes of the Late-Time Universe Kunhao Zhong, Bhuvnesh Jain Abstract Recent results from the Dark Energy Spectroscopic Instrument (DESI) have shown a strong statistical preference for a time-evolving dark energy model over Λ \Lambda CDM when combining BAO, CMB, and supernova (SN) data. We investigate the robustness of this conclusion by isolating geometric information in weak lensing measurements from the DES Year 3 survey and combining it with different datasets. We introduce a hyperparameter, Ω m growth \Omega_{\rm m}^{\rm growth} , to decouple the growth contribution from the lensing 2-point correlation and thus bypass the possible effect of the σ 8 \sigma_{8} tension in our analysis. We then combine with the late-time geometric probes provided by BAO and SN, along with CMB primary data. The preference for evolving dark energy is consistent with the DESI-DR2 findings: when combining BAO, primary CMB, and weak lensing data, the w 0 w a w_{0}w_{a} CDM is preferred at about the 3 σ 3\sigma significance. However, when we add SN, the result is sensitive to the choice of data: if we leave out z < 0.1 z<0.1 SN data in the analysis, as a test of the effect of inhomogeneous calibration, we obtain a statistical significance below 2 σ 2\sigma for time evolving dark energy. Indeed, the high-z only SN data lowers the evidence for evolving dark energy in the data combinations we have examined. This underscores the importance of improved SN samples at low redshift and of alternative data combinations. We show that cosmic shear measurements with LSST Year 1 data will provide comparable power to current SN data. We discuss other low-redshift probes provided by lensing and galaxy clustering to test for evolving dark energy. I Introduction Recent results from the Dark Energy Spectroscopic Instrument (DESI) and type Ia supernova (SN) challenge the \Lambda CDM model, showing a strong statistical preference for time-evolving dark energy Adame et al. 2025 ; Abdul Karim et al. 2025 ; Lodha et al. 2025 ; Abbott et al. 2025 . A joint analysis of baryon acoustic oscillation (BAO), cosmic microwave background (CMB), and SN data prefers the w_{0}w_{a} CDM model over \Lambda CDM by up to 4.2\sigma . The highest significance, 4.2\sigma , is achieved when combining DESI data with external supernova data from the Dark Energy Survey Year 5 (DES-Y5) and CMB measurements, including the full temperature and polarization acoustic peaks as well as CMB lensing. The DESI Data Release 2 Abdul Karim et al. 2025 (hereafter DESI-DR2) includes a series of robustness tests to verify this potentially groundbreaking result. One such test examines whether the preference for dynamical dark energy is independent of the early-universe CMB constraint on the matter density \Omega_{\rm m} . Replacing the CMB data with low-redshift weak lensing measurements from the 3-year Dark Energy Survey (DES-Y3), they found consistent results. However, DES-Y3 data combined with BAO shows only a modest preference for dynamical dark energy ( 2.2\sigma ) and does not increase the overall significance when supernova data is included. In contrast to distance measurements (geometry), constraining structure formation through perturbations (growth) is significantly more challenging. A long-standing debate centers on whether direct probes of structure growth yield a lower clustering amplitude than that predicted by the CMB, a discrepancy known as the \sigma_{8} tension Troxel et al. 2018 ; Abbott et al. 2018 ; Abbott et al. 2022 ; Dalal et al. 2023 ; Asgari et al. 2021 ; Stölzner et al. 2025 . This difficulty is not unique to weak lensing; it also affects analyses based on redshift-space distortions (RSD), full-shape power spectrum, and other low-redshift measurements. See Ref. Abdalla et al. 2022 for a review. Depending on the analysis method and dataset, the inferred lensing amplitude from DESI data varies Sailer et al. 2025 ; Chen et al. 2024a ; Adame et al. 2024a ; Adame et al. 2024b ; Maus et al. 2025 . There is a mild tension with results from BOSS Chen et al. 2024b ; Chen et al. 2022 . The situation becomes even more complex when the expansion history is allowed to vary within the w_{0}w_{a} framework. Nevertheless, with the inclusion of SN data, most analyses still report a consistent preference for the w_{0}w_{a} model Chen et al. 2024b ; Karim et al. 2025 ; Maus et al. 2025 . Conversely, whether the dynamical dark energy evolution can help solve the S8 tension is an interesting but still open question Heydenreich et al. 2025 . In this work, we aim to simplify the problem by isolating the geometric information from weak lensing probes. By marginalizing over the growth contribution, we can robustly combine weak lensing and BAO data to test whether the preference for dynamical dark energy persists. Our methodology largely follows previous approaches that assess the consistency between geometry and growth Muir et al. 2021 ; Ruiz-Zapatero et al. 2021a ; Zhong et al. 2023 . When testing the w_{0}w_{a} model, we treat the growth information as a nuisance component to marginalize over, rather than attempting to extract precise constraints on the lensing amplitude. By inflating the uncertainty on \sigma_{8} , we test whether the statistical significance of the w_{0}w_{a} preference is affected. The main constraining power in this approach arises from the ratios of different two-point statistics. The structure of this paper is as follows. In Sec. II we introduce how we marginalize growth information and other choices different from DESI-DR2, including the effort to better estimate tension. We present results in Sec. III and conclude in Sec. IV . II Methodology II.1 Growth Marginalized Weak Lensing We introduce the growth-geometry split in this section. The analysis mainly follows the previous analysis in Ref Muir et al. 2021 ; Zhong et al. 2023 ; Ruiz and Huterer 2015 ; Bernal et al. 2016 ; Andrade et al. 2021 ; Ruiz-Zapatero et al. 2021b . The goal of this work is not to extract precise information about structure growth, but rather to marginalize over its contribution in order to isolate the geometric component, which can then be combined with BAO and SN data. Thus, we expect the specific implementation choices to have minimal impact on the main results. To split growth and geometry, we introduce a hyperparameter \Omega_{\rm m}^{\rm growth} . Since the late-time linear matter power spectrum can be written as the product of the primordial power spectrum, the transfer function, and the growth function G_{\rm growth} , we define the split power spectrum as follows: \displaystyle P_{\rm split}(k,z) \displaystyle=P_{\rm geo}^{\rm linear}(k,z)\times\bigg(\dfrac{G_{\rm growth}^{\rm ODE}(z)}{G_{\rm geo}^{\rm ODE}(z)}\bigg)^{2}\times B^{\rm growth}(k,z). (1) Here, the growth function G(z) describes the growth of structure perturbations. \times B^{\rm growth}(k,z) is the boost factor that corrects for non-linear effects. G(z) is a function of parameters \Omega_{\rm m} , w_{0} , and w_{a} : G^{\prime\prime}+\left(4+\frac{H^{\prime}}{H}\right)G^{\prime}+\left[3+\frac{H^{\prime}}{H}-\frac{3}{2}\frac{\Omega_{\mathrm{m}}^{0}a^{-3}}{E^{2}(a)}\left(1-f_{v}\right)\right]G=0, (2) where in w_{0}w_{a} CDM: \displaystyle E^{2}(a) \displaystyle\equiv H(a)^{2}/H_{0}^{2} (3) \displaystyle=\Omega_{\mathrm{m}}^{0}a^{-3}+\Omega_{\mathrm{r}}^{0}a^{-4}+\Omega_{\mathrm{\Lambda}}^{0}a^{-3\left(1+w_{0}+w_{\mathrm{a}}\right)}e^{-3w_{\mathrm{a}}(1-a)}. We assume a flat universe \Omega_{\mathrm{\Lambda}}^{0}=1-\Omega_{\mathrm{m}}^{0} . Neutrino effect is accounted for through the fractional mass term f_{\nu}=\Omega_{\nu}/\Omega_{\mathrm{m}} . Prime in eq. 2 denotes derivative with respect to \log a . Eq. 2 is a function of background parameters. Thus, by redefining the power spectrum as eq. 1 , we separated the