Professional Documents
Culture Documents
measurements from [2], H0 = 74.03 ± 1.42 km/s/Mpc. This failure of the generalized GPT scenario to host
the P18-BAO-R19 trio is of significance and noticeably shrinks the space of possible gravitational solutions to
the Hubble tension.
I. INTRODUCTION AND MOTIVATIONS In a recent work [1] the authors proposed a phenomeno-
logical phase transition in the gravity sector by assuming that
The recent cosmological tensions between early and late the early standard ΛCDM transits to a modified gravity model
time observations in the context of the standard model of both at the background and perturbation levels. We verified
cosmology, the so-called ΛCDM, have been extensively ad- that the model could resolve the CMB-lensing, low- and high-
dressed by many cosmologists in the last few years. However, `’s, σ8 and H0 tensions simultaneously (at different levels).
a truly convincing theoretical model for their explanation is The BAO datasets were not included in the main analysis and
still missing. The main tension is the well-known H0 -tension parameter estimation of [1]. However, the predictions of GPT
which is the ∼ 4.2σ discrepancy between the local measure- for BAO observables were shown to disagree with some BAO
ments of the Hubble constant [2, 3, 4] and its inferred value data points. In this work we generalize the model to allow for
from the cosmic microwave background (CMB) data in the transition between two general early and late gravity models
ΛCDM framework [5]1 . There have also been many attempts at some intermediate redshift. We will study this model in the
for H0 estimation from other independent observations, with presence of BAO datasets. These standard rulers provide us
higher H0 values [8, 9, 10, 11, 12]. Beside the Hubble tension, with independent observations at low redshifts, where a tran-
there is the milder σ8 tension with its local measurements im- sition in the gravity sector was shown to relax the tensions
pling less clumpiness in matter distribution compared to the between CMB and local Hubble measurements. On one hand,
CMB-based inferences [5]. Certain internal inconsistencies BAO data are expected to break the parameter degeneracies
are also reported in Planck results, specifically in the CMB and improve parameter measurements. On the other hand, the
lensing amplitude and the low- and high-` discrepancy. If enhanced space of gravitational degrees of freedom at early
not due to systematics, and if considered statistically signif- times may have the potential to accommodate both BAO data
icant, theoretical models beyond ΛCDM are required to ex- and local H0 values. This is the driving motivation of this pa-
plain them. per: to explore the huge subset of gravity models effectively
modelled by a transition at some redshift, with possible si-
Many different models are proposed to address these cos- multaneous deviations from GR in early and late times, and
mological tensions [see 13, for a comprehensive review]. The search for a possible sweet spot where CMB, BAO and local
models could in general be categorized into late and early, H0 measurements can all co-exist in peace.
based on the era of the modification. The motivation behind
the late time modifications is clear and there are many at-
tempts [see, e.g., 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, II. THE GPT MODEL
25, 26, 27, 28, 29, 30, 31]. The early modifications try to
reduce the sound horizon at the last scattering surface which
results in a higher value for H0 [32, 33, 34, 35]. There are The gravitational phase transition, or GPT model, intro-
many other different ideas to approach the H0 tension, such duced in [1], suggests a transition in the gravitational theory,
as dark matter [36], primordial magnetic fields [37] and etc. phenomenologically modelled by a "tanh" function. It was
Neither of these models are consistent with all the datasets, in assumed that the early phase of gravitation is governed by
particular when data from the baryonic acoustic oscillations the standard Einstein-Hilbert model, referred to as GR, which
(BAO) are included [38]. forms the gravity model in the ΛCDM scenario. In this work
we relax this assumption and study the consequences.
We assume our Universe is isotropic, homogeneous and
flat, where the background evolution is given by
∗ Electronic address: n-khosravi@sbu.ac.ir
† Electronicaddress: m farhang@sbu.ac.ir 2 2 4 3
1 H (z) = H0 Ωr (1 + z) + Ωm (1 + z) + ΩΛ (z) (1)
For some controversy on the significance of the Hubble tension see [6, 7].
2
P
as
k 2 Ψ = −µ(z) 4πGa2 [ρ∆ + 3(ρ + P )σ] (3)
k 2 [Φ − γ(z)Ψ] = µ(z) 12πGa2 (ρ + P )σ, (4)
where µ(z) and γ(z) are responsible for any general redshift-
dependent modifications to GR. Equations (3) and (4) can also
be written as 50 60 70 80 90 100
k 2 (Φ + Ψ) = 8πGa2 Σ(z) ρ ∆ (5)
H0
which is particularly useful for studying the gravitational lens- FIG. 1: The 1-D marginalized likelihood for the Hubble constant
ing. In the GPT scenario, we assume a transition in the gravity H0 is shown for different datasets. The P18 case, represented by
sector and model the transition through "tanh" function. At the solid green curve, shows no inconsistencies with R19. Adding
the background level we therefore have R19 (the dashed red curve) results in a shift to higher H0 values as
expected. In P18+BAO case, however, the preferred H0 value, repre-
early 1 + tanh α (zt − z) sented by dotted blue curve, becomes much smaller in comparison to
ΩΛ (z) = ΩΛ + ∆Λ . (6) R19. This shows the power of BAO in putting very tight constraints
2
on our model parameters.
Here zt is the transition redshift and α is the inverse of the
transition’s width. Imposing the flatness condition gives Ωr +
Ωm + ΩΛ = 1, where ΩΛ ≡ ΩΛ (z = 0) = Ωearly + ∆Λ [1 + dataset consists of 6DFGS [39], SDSS MGS [40] and BOSS
tanh(αzt )]/2. Similarly, the modifications to the perturbation DR12 [41] and we refer to them in general as BAO.
equations are characterized by transitions in µ(z) and γ(z), We use CosmoMC3 , the Cosmological Monte
Carlo code [42], to estimate the gGPT parameters
early 1 + tanh α (zt − z) (∆Λ , µearly , ∆µ , γ early , ∆γ , α, zt ) as well as the stan-
µ(z) = µ + ∆µ ,
2 dard cosmological parameters (Ωb h2 , Ωdm h2 , θ, τ, ns , As )
1 + tanh α (zt − z) with the above datasets. We assume uniform prior on all of
γ(z) = γ early + ∆γ . (7) these parameters. For parameter estimation we use P18+BAO
2
and P18+R19. We know the consistency of these data
where we do not impose any restricting assumptions on µearly , combinations from the standard ΛCDM and the late GPT
γ early and Σearly . This is in contrast to the late GPT model model respectively. To be more precise, there is no tension
where we had µearly = γ early = 1 and Σearly = −1. This between P18 and BAO in ΛCDM and between P19 and R19
generalization allows for transitions happening between two in late GPT. Therefore there should be no tension in the
general gravity models and not necessarily starting from GR extended encompassing parameter space of gGPT neither.
in the past. We refer to this generalized model as gGPT.
Ω b h2 Ω c h2 H0 τ log(1010 As ) ns
P18+R19 0.0227 ± 0.0002 0.1182 ± 0.0016 73.69 ± 1.47 0.049 ± 0.008 3.03 ± 0.02 0.968 ± 0.005
P18+BAO 0.0226 ± 0.0002 0.1183 ± 0.0012 68.03 ± 0.76 0.050 ± 0.008 3.03 ± 0.02 0.967 ± 0.004
TABLE I: Standard cosmological parameters as measured by the data combinations, P18+R19 and P18+BAO, in the gGPT framework.
∆H (z)
found to be consistent with R19 [1]. However, we find that
0
adding BAO to P18 data would tighten the likelihood around
the lower value of H0 = 68.03 ± 0.76, and thus disagree −2
with R19 measurements. We should emphasize that these re-
−4
sults are in the presence of the early-, as well as late-time,
degrees of freedom, allowing for the gravity theory to devi- −6
ate from GR even before transitioning to a possibly modified
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
late gravity model. The other standard parameters, measured
z
with P18+BAO and P18+R19, are found to be consistent, as
presented in Table I. P18
P18+R19
In the following we investigate how different data combi- 1.0 P18+BAO
nations would constrain the evolution of the main free param-
eters in the model.
Background evolution: Fig. 2 shows the evolution of H 0.8
ΩΛ(z)
and ΩΛ , calculated for various data combinations. For better
visual comparison, the P18 best-fit measurement of H(z) is 0.6
V. FINAL WORDS
4 It should be noted that, although the x-axis in the ∆H (z) plot, by defini- In this work we generalized our previous GPT model [1]
tion, corresponds to the best-fit gGPT with P18 data, its agreement with
the best-fit ΛCDM is quite well (as it is obvious from solid green line in to allow the gravity sector go beyond GR both in early and
ΩΛ (z) plot). However, the 1σ spans of the trajectories around this com- late times. In [1] we showed that a late-only phase of de-
mon best-fit differ by a large amount. viation from GR can address H0 and σ8 tension, as well as
4
1.30
P18 the internal inconsistencies of Planck. However, there were
P18+R19
P18+BAO
hints for tensions with some BAO observations based on GPT
1.25
predictions for BAO distances. In the current work we in-
1.20 cluded BAO in our study of the gGPT scenario to fully study
1.15
their impact on parameter estimation. In brief, even in this en-
hanced framework, BAO did not allow for deviations from GR
µ(z)
1.10
in the background level. Therefore, our gGPT model could
1.05 not resolve the H0 tension among the P18-R19-BAO trio. At
1.00 the perturbation level, on the other hand, the data are found
to be consistent. The ∼ 1σ deviation of µ and γ from GR
0.95
predictions with BAO, although not statistically significant, is
0.90 persuading for future investigation of these perturbation pa-
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
rameters with more informative data and hints to the power
z
of these parameters in differentiating between various scenar-
P18 ios and better understanding of the origin of the cosmological
P18+R19
1.4
P18+BAO
tensions including H0 and σ8 .
The failure of this relatively general and effective extension
1.2 of GR, both in the background and in perturbations, should
be taken seriously. That is because the proposed gGPT sce-
γ(z)
−0.92
−0.94
Acknowledgement
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
z
The numerical calculations of this work were carried out on
FIG. 4: The correlation between µ(z) and γ(z), for the three data the computing cluster of the Canadian Institute for Theoretical
combinations used in this work. Astrophysics (CITA), University of Toronto.
[1] M. Farhang and N. Khosravi, Phys. Rev. D 103, 083523 (2021), [4] A. G. Riess et al., Astrophys. J. 861, 126 (2018), 1804.10655.
2011.08050. [5] Planck Collaboration, N. Aghanim, Y. Akrami, M. Ashdown,
[2] A. G. Riess, S. Casertano, W. Yuan, L. M. Macri, and D. Scol- J. Aumont, C. Baccigalupi, M. Ballardini, A. J. Banday, R. B.
nic, Astrophys. J. 876, 85 (2019), 1903.07603. Barreiro, N. Bartolo, et al., Astron. Astrophys. 641, A6 (2020),
[3] A. G. Riess et al., Astrophys. J. 826, 56 (2016), 1604.01424. 1807.06209.
5
[6] G. Efstathiou, Mon. Not. Roy. Astron. Soc. 505, 3866 (2021), [26] A. Gómez-Valent, V. Pettorino, and L. Amendola, Phys. Rev. D
2103.08723. 101, 123513 (2020), 2004.00610.
[7] D. Camarena and V. Marra, Mon. Not. Roy. Astron. Soc. 504, [27] M. Lucca and D. C. Hooper (2020), 2002.06127.
5164 (2021), 2101.08641. [28] E. Di Valentino, A. Melchiorri, O. Mena, and S. Vagnozzi,
[8] W. L. Freedman et al. (2019), 1907.05922. Phys. Dark Univ. 30, 100666 (2020), 1908.04281.
[9] C. D. Huang, A. G. Riess, W. Yuan, L. M. Macri, N. L. Za- [29] E. Di Valentino, A. Melchiorri, O. Mena, and S. Vagnozzi,
kamska, S. Casertano, P. A. Whitelock, S. L. Hoffmann, A. V. Phys. Rev. D 101, 063502 (2020), 1910.09853.
Filippenko, and D. Scolnic (2019), 1908.10883. [30] A. Banihashemi, N. Khosravi, and A. H. Shirazi, Phys. Rev. D
[10] K. C. Wong et al., Mon. Not. Roy. Astron. Soc. 498, 1420 101, 123521 (2020), 1808.02472.
(2020), 1907.04869. [31] A. Banihashemi, N. Khosravi, and A. Shafieloo, JCAP 06, 003
[11] N. Khetan et al. (2020), 2008.07754. (2021), 2012.01407.
[12] D. Pesce et al., Astrophys. J. Lett. 891, L1 (2020), 2001.09213. [32] V. Poulin, T. L. Smith, T. Karwal, and M. Kamionkowski, Phys.
[13] E. Di Valentino, O. Mena, S. Pan, L. Visinelli, W. Yang, A. Mel- Rev. Lett. 122, 221301 (2019), 1811.04083.
chiorri, D. F. Mota, A. G. Riess, and J. Silk, Class. Quant. Grav. [33] L. Knox and M. Millea, Phys. Rev. D 101, 043533 (2020),
38, 153001 (2021), 2103.01183. 1908.03663.
[14] E. Di Valentino, A. Melchiorri, and O. Mena, Phys. Rev. D 96, [34] R. E. Keeley, A. Shafieloo, D. K. Hazra, and T. Souradeep,
043503 (2017), 1704.08342. JCAP 09, 055 (2020), 2006.12710.
[15] G.-B. Zhao et al., Nature Astron. 1, 627 (2017), 1701.08165. [35] S. Vagnozzi (2021), 2105.10425.
[16] N. Khosravi, S. Baghram, N. Afshordi, and N. Altamirano, [36] K. Vattis, S. M. Koushiappas, and A. Loeb, Phys. Rev. D 99,
Phys. Rev. D 99, 103526 (2019), 1710.09366. 121302 (2019), 1903.06220.
[17] W. Yang, S. Pan, E. Di Valentino, R. C. Nunes, S. Vagnozzi, [37] K. Jedamzik and L. Pogosian, Phys. Rev. Lett. 125, 181302
and D. F. Mota, JCAP 09, 019 (2018), 1805.08252. (2020), 2004.09487.
[18] W. Yang, A. Mukherjee, E. Di Valentino, and S. Pan, Phys. Rev. [38] K. Jedamzik, L. Pogosian, and G.-B. Zhao, Commun. in Phys.
D 98, 123527 (2018), 1809.06883. 4, 123 (2021), 2010.04158.
[19] A. Banihashemi, N. Khosravi, and A. H. Shirazi, Phys. Rev. D [39] F. Beutler, C. Blake, M. Colless, D. H. Jones, L. Staveley-
99, 083509 (2019), 1810.11007. Smith, L. Campbell, Q. Parker, W. Saunders, and F. Watson,
[20] X. Li and A. Shafieloo, Astrophys. J. Lett. 883, L3 (2019), Mon. Not. Roy. Astron. Soc. 416, 3017 (2011), 1106.3366.
1906.08275. [40] A. J. Ross, L. Samushia, C. Howlett, W. J. Percival, A. Burden,
[21] R. E. Keeley, S. Joudaki, M. Kaplinghat, and D. Kirkby, JCAP and M. Manera, Mon. Not. Roy. Astron. Soc. 449, 835 (2015),
12, 035 (2019), 1905.10198. 1409.3242.
[22] M. Raveri, Phys. Rev. D 101, 083524 (2020), 1902.01366. [41] S. Alam et al. (BOSS), Mon. Not. Roy. Astron. Soc. 470, 2617
[23] S.-F. Yan, P. Zhang, J.-W. Chen, X.-Z. Zhang, Y.-F. Cai, and (2017), 1607.03155.
E. N. Saridakis, Phys. Rev. D 101, 121301 (2020), 1909.06388. [42] A. Lewis and S. Bridle, Phys. Rev. D 66, 103511 (2002), astro-
[24] E. Di Valentino, A. Mukherjee, and A. A. Sen (2020), ph/0205436.
2005.12587. [43] L. Pogosian, M. Raveri, K. Koyama, M. Martinelli, A. Silvestri,
[25] M. Braglia, M. Ballardini, W. T. Emond, F. Finelli, A. E. Gum- and G.-B. Zhao (2021), 2107.12992.
rukcuoglu, K. Koyama, and D. Paoletti, Phys. Rev. D 102,
023529 (2020), 2004.11161.