Periodic Reporting for period 1 - NewRecords (New crisis and old mysteries: Resolving cosmic tensions to reveal the dark sector)
Reporting period: 2023-01-01 to 2025-06-30
It has recently been established that the current expansion rate of the universe (i.e the Hubble constant) as measured by probes of the ``late-universe'' is discrepant at 5σ with that predicted in the standard Λ cold dark matter (ΛCDM) model calibrated onto cosmic microwave background (CMB) data. Additionally, analyses of the latest galaxy weak lensing surveys find an amplitude of density fluctuations on large scales (the S8 parameter) that is about 2−3σ lower than what is predicted in ΛCDM. This growing “crisis” is potentially the first indication of a departure from the ΛCDM model, giving hopes to access the fundamental properties of the mysterious dark matter and dark energy that pervade the universe. On the one hand, I have shown with my colleagues that the “Hubble tension” could indicate the presence of new physics in the era pre-recombination, potentially connected to new dark energy properties at early time. On the other hand, the “S8 tension” can be connected to new dark matter properties, and in particular could indicate that dark matter is not stable on cosmological timescales, or interacts with some new light species. Yet, none of the suggested models have been robustly detected nor are they able to explain simultaneously the Hubble and S8 tensions. With the support of the ERC, I will build a group that will develop public tools to firmly identify the new physics mechanism required to explain these cosmic tensions, assess the observational consequences for a wide variety of observables, and the implications for the nature of dark matter and dark energy.
The activities of the group are devised between three main work packages, following the objectives set in the Description of the Action:
Work Package 1: Signatures in the CMB
We performed updated analyses of promising models to resolve cosmic tensions using the latest CMB data from Planck, ACT, and SPT. Our focus was on Early Dark Energy (EDE), early modified gravity, and modified recombination models. To enable these studies, we significantly improved the public code CLASS, allowing us to compute the effects of previously unexplored models. These enhancements included implementing a dark energy–dark matter drag (and its extension to EDE-DM interactions), generalizing initial conditions for the EDE field beyond slow roll, introducing a pre-recombination modified gravity model, and adopting a model-independent method to reconstruct the recombination history. We also developed a new frequentist analysis tool to complement Bayesian methods and contributed updated Planck CMB likelihoods in MontePython. These technical advances led to 7 publications and multiple code releases.
Work Package 2: Signatures in the Large-Scale Structures
Our work here involved computing novel signatures of promising models for cosmic tensions in large-scale structure data, including galaxy and quasar surveys (eBOSS and DESI), the lyman-α forest, and 21cm surveys (HERA). We have compared different formalisms of the effective field theory of large-scale structures (EFTofLSS) applied to BOSS and eBOSS data, which led to different constraints on LCDM and its extensions, using a Frequentist analysis framework to contrast Bayesian analysis results. We also carried out N-body simulations to assess the impact of primordial non-gaussianity on the S8 tension and developed a simulation pipeline covering scales from the cosmological horizon to halo levels. Additionally, we derived new constraints from lyman-alpha data, made forecasts for future 21cm surveys, and implemented a new likelihood of the recent DESI data in MontePython, resulting in 6 publications and further code releases.
Work Package 3: Theoretical Implications for the Dark Sector
This package focused on the implications of EDE models for understanding Dark Energy and related cosmological issues such as the age of the universe and inflationary properties. We produced a comprehensive review comparing current constraints on various EDE models, highlighting their general implications and challenges. Expanding beyond EDE, we examined models that modify the pre-recombination era to resolve the Hubble tension and explored synergies between early and late-time modifications for guiding model building. We advanced standard axion-like EDE models through revised initial conditions, non-minimal gravity couplings, and dark matter interactions, contributing to several CMB-related papers. Moreover, we developed an extended “New” Early Dark Energy model featuring a first-order phase transition and contributed to a bi-metric theory of gravity addressing the curvature tension between CMB and BAO data, all disseminated via multiple peer-reviewed publications.
Work Package 1: Signatures in the CMB
We performed updated analyses of promising models to resolve cosmic tensions using the latest CMB data from Planck, ACT, and SPT. Our focus was on Early Dark Energy (EDE), early modified gravity, and modified recombination models. To enable these studies, we significantly improved the public code CLASS, allowing us to compute the effects of previously unexplored models. These enhancements included implementing a dark energy–dark matter drag (and its extension to EDE-DM interactions), generalizing initial conditions for the EDE field beyond slow roll, introducing a pre-recombination modified gravity model, and adopting a model-independent method to reconstruct the recombination history. We also developed a new frequentist analysis tool to complement Bayesian methods and contributed updated Planck CMB likelihoods in MontePython. These technical advances led to 7 publications and multiple code releases.
Work Package 2: Signatures in the Large-Scale Structures
Our work here involved computing novel signatures of promising models for cosmic tensions in large-scale structure data, including galaxy and quasar surveys (eBOSS and DESI), the lyman-α forest, and 21cm surveys (HERA). We have compared different formalisms of the effective field theory of large-scale structures (EFTofLSS) applied to BOSS and eBOSS data, which led to different constraints on LCDM and its extensions, using a Frequentist analysis framework to contrast Bayesian analysis results. We also carried out N-body simulations to assess the impact of primordial non-gaussianity on the S8 tension and developed a simulation pipeline covering scales from the cosmological horizon to halo levels. Additionally, we derived new constraints from lyman-alpha data, made forecasts for future 21cm surveys, and implemented a new likelihood of the recent DESI data in MontePython, resulting in 6 publications and further code releases.
Work Package 3: Theoretical Implications for the Dark Sector
This package focused on the implications of EDE models for understanding Dark Energy and related cosmological issues such as the age of the universe and inflationary properties. We produced a comprehensive review comparing current constraints on various EDE models, highlighting their general implications and challenges. Expanding beyond EDE, we examined models that modify the pre-recombination era to resolve the Hubble tension and explored synergies between early and late-time modifications for guiding model building. We advanced standard axion-like EDE models through revised initial conditions, non-minimal gravity couplings, and dark matter interactions, contributing to several CMB-related papers. Moreover, we developed an extended “New” Early Dark Energy model featuring a first-order phase transition and contributed to a bi-metric theory of gravity addressing the curvature tension between CMB and BAO data, all disseminated via multiple peer-reviewed publications.
All the following items represent new results, beyond the state of the art in the field:
-- In a couple of papers, we have established that the preference seen by ACT DR4 in favor of the `axion-like'' EDE model over LCDM was not supported by up-to-date Planck and SPT data. This has led to a publication in Physical Review Letters. Future ACT data will firmly establish the fate of the axion-like'' EDE model.
-- Through a novel model-independent method, we established the limitations of non-standard recombination models in addressing cosmic tensions, leading to a publication in Physical Review Letters.
-- We identified a new statistically significant tension between lyman-alpha forest data and Planck, which—if not due to unresolved systematic errors—could hint at new physics. This tension might be explained by models of warm dark matter, ultra-light axion dark matter, or a non-standard primordial power spectrum. Upcoming DESI survey data, coupled with improved modeling, will help distinguish the source.
-- Using the Fisher formalism, we demonstrated that future 21cm data from HERA could improve constraints on models of DM-DR interactions by up to four orders of magnitude when the interaction is strong enough to explain the S8 tension. We now await real data, alongside working on improvements in the speed of the 21cm power spectrum computations.
-- We found that the recent hint of exotic neutrino physics is tied to a known A_L anomaly in CMB data and a single BAO data point in tension with earlier results. This hint vanishes with updated CMB data lacking the A_L anomaly and after removing the anomalous BAO point.
-- We showed that modifications to standard axion-like EDE models, such as adopting different initial conditions (beyond slow-roll), coupling to dark matter, or incorporating non-minimal coupling to gravity, can improve model performance. These modifications reduce S8 and revive models previously excluded from addressing the H0 tension. Next steps include testing this model with updated LSS data and refining theoretical predictions.
-- Dedicated N-body simulations revealed that primordial non-gaussianity may help resolve the S8 tension. It remains to be explored whether combining EDE (or other mechanisms resolving the H0 tension) with primordial non-gaussianities can jointly resolve both tensions and yield predictions for future surveys.
A total of 15 articles were published in peer-reviewed journals over the past two years and have already gathered close to 500 citations, including a review paper with overs 150 citations. This illustrates the high relevance of the work performed in the group for the community, and the very intense activity around these topics.
-- In a couple of papers, we have established that the preference seen by ACT DR4 in favor of the `axion-like'' EDE model over LCDM was not supported by up-to-date Planck and SPT data. This has led to a publication in Physical Review Letters. Future ACT data will firmly establish the fate of the axion-like'' EDE model.
-- Through a novel model-independent method, we established the limitations of non-standard recombination models in addressing cosmic tensions, leading to a publication in Physical Review Letters.
-- We identified a new statistically significant tension between lyman-alpha forest data and Planck, which—if not due to unresolved systematic errors—could hint at new physics. This tension might be explained by models of warm dark matter, ultra-light axion dark matter, or a non-standard primordial power spectrum. Upcoming DESI survey data, coupled with improved modeling, will help distinguish the source.
-- Using the Fisher formalism, we demonstrated that future 21cm data from HERA could improve constraints on models of DM-DR interactions by up to four orders of magnitude when the interaction is strong enough to explain the S8 tension. We now await real data, alongside working on improvements in the speed of the 21cm power spectrum computations.
-- We found that the recent hint of exotic neutrino physics is tied to a known A_L anomaly in CMB data and a single BAO data point in tension with earlier results. This hint vanishes with updated CMB data lacking the A_L anomaly and after removing the anomalous BAO point.
-- We showed that modifications to standard axion-like EDE models, such as adopting different initial conditions (beyond slow-roll), coupling to dark matter, or incorporating non-minimal coupling to gravity, can improve model performance. These modifications reduce S8 and revive models previously excluded from addressing the H0 tension. Next steps include testing this model with updated LSS data and refining theoretical predictions.
-- Dedicated N-body simulations revealed that primordial non-gaussianity may help resolve the S8 tension. It remains to be explored whether combining EDE (or other mechanisms resolving the H0 tension) with primordial non-gaussianities can jointly resolve both tensions and yield predictions for future surveys.
A total of 15 articles were published in peer-reviewed journals over the past two years and have already gathered close to 500 citations, including a review paper with overs 150 citations. This illustrates the high relevance of the work performed in the group for the community, and the very intense activity around these topics.