Impacts of Dynamical Dark Energy on the Neutrino Mass Constraints

Gabriel Rodrigues,1, ∗ Jamerson Rodrigues,1, † and Jailson Alcaniz1, ‡

1
Observatório Nacional, Rio de Janeiro - RJ, 20921-400, Brasil
(Dated: March 4, 2025)
P
The difference between the total neutrino mass estimates ( mν ) derived from cosmological data
within the standard ΛCDM model and those obtained from terrestrial particle physics experiments
underscores the need to explore alternative scenarios. Recent analyses have shown that a dynamic
dark energy modeled byP the CPL parameterization of the dark energy equation of state (EoS) can
ease the constraints on mν , thus alleviating this tension. This study investigates the robustness
of this discrepancy by assessing the extent to which the CPL assumption influences the results.
We examine how other EoS parameterizations, P such as the Barboza-Alcaniz (BA) and Jassal-Bagla-
arXiv:2503.00126v1 [astro-ph.CO] 28 Feb 2025

Padmanabhan (JBP) parameterizations, affect mν estimates. We perform a Markov Chain Monte

Carlo (MCMC) analysis combining the latest baryon acoustic oscillation data from DESI with
the Planck 2018 cosmic microwave background data, which includes information on temperature,
polarization, and lensing, as well as the Pantheon+ Type Ia supernovae observations. While both
the BA and JBP parameterizations can also resolve the tension, our results show a correlation
between the dark energy EoS and the constraints on neutrino mass.

I. INTRODUCTION such models tightly. This limitation is due to a geometric

degeneracy between the parameters that affect the late-
In the standard cosmology, a dark energy (DE) com- time Cosmology and the angular diameter distance to
ponent drives the current cosmic acceleration. This com- the surface of the last scattering. Therefore, this degen-
ponent possesses negative pressure and is characterized eracy must be broken using local Universe observables
by an Equation of State (EoS) parameter defined as to constrain this class of models. Although incorporat-
w = p/ρ, where p is the dark energy pressure and ρ its en- ing information such as SN data alongside CMB obser-
ergy density. A natural theoretical representation of DE vations helps tighten the constraints, no preference for
is the vacuum energy density, also known as the cosmo- this class of models has been shown at first. This re-
logical constant Λ, for which w = −1. However, despite mained the case until very recently, when BAO obser-
being able to explain most of the present-day observa- vations from the Dark Energy Spectroscopic Instrument
tional data, the standard Λ-Cold Dark Matter (ΛCDM) (DESI), combined with Planck CMB and Pantheon+ SN
model faces tensions in measurements of some cosmolog- data, showed a significant preference for DDE using the
ical parameters. The most prominent of these discrepan- Chevallier-Polarski-Linder (CPL) EoS parameterization,
cies is a tension of ∼ 5σ between the Hubble Constant H0 w(a) = w0 + wa (1 − a) [24]. Even more recently, it has
measurements in the late- and early-time Universe. Fur- been shown that this preference is not limited to the CPL
thermore, Λ is associated with well-known fundamental but extends to other parameterizations as well. This re-
problems in theoretical physics [1–4]. sult further reinforces the preference indicated by para-
In this context, various alternative proposals have been metric analyses of the DESI and SNe data for DDE mod-
suggested to explain the current cosmic acceleration be- els [25]1 .
yond the cosmological constant (see e.g. [5–11]). In gen- Beyond the DDE discussion, the late-time universe
eral, such proposals assume the existence of scalar fields is plagued by further uncertainties related to the back-
that have not yet reached their ground state or mod- ground and perturbations evolution, and the neutrino
ifications in the Friedmann equations, implying time- physics is of particular interest in this picture. In the
dependent EoS parameters w(a). Phenomenologically, standard scenario, outlined by the Standard Model of
several authors have also proposed w(a) parameteriza- fundamental particles (SM), neutrinos are massless par-
tion [12–16], which insert the ΛCDM model in a more ticles. Such a conclusion is challenged by the num-
general framework, opening the possibility of checking ber density measurements of the solar and atmospheric
deviations from the standard case. neutrinos, pointing to an oscillatory behavior of the
To assess the observational viability of dynamical dark three flavors. Oscillation experiments are insensitive
energy (DDE) models, several probes have been utilized to the absolute mass scale of the neutrinos, but use-
from both the early and local Universe [17–23]. It is im- ful to probe their squared mass differences. The latter
portant to note that CMB data alone cannot constrain results from atmospheric neutrinos point to |∆m232 | ≈
2.4 × 10−3 eV2 , while from solar neutrinos one obtains

∗ [email protected] 1 This preference for a DDE is not confirmed by non-parametric

† [email protected] (model-independent) analyses of the current BAO and SN data,
‡ [email protected] as shown in [26, 27] (see also [28]).
 2

|∆m221 | ≈ 7.4 × 10−5 eV2 [29]. This configuration en- the fluid velocity. If we assume a Newtonian gauge,
ables at least two non-relativistic species of neutrinos.
Lower limits ds2 = −(1 + 2Ψ)dt2 + a2 (t)(1 − 2Φ)δij dxi dxj , (2)
P on the sum of their masses can also be es-
timated: mν > 0.06 eV for the Normal Hierarchy
schemes (NH), while the Inverted Hierarchy (IH) requires then, the Continuity and Euler equations are written as
P
mν > 0.10 eV.  
Back to the cosmological perspective, neutrinos have a δ̇ + 3H(c2s − w)δ + (1 + w) θ − 3Φ̇ = 0, (3a)
well-marked role in the expansion of the cosmos. They
act as radiation in the very early universe, later changing
k 2 c2s δ
to a hot dark matter behavior after the non-relativistic θ̇ + H(1 − 3w)θ − − k 2 Ψ = 0. (3b)
transition. At the perturbative level, the neutrinos dis- 1+w
persion acts to suppress the growth of perturbation at The dark energy also affects the adiabatic sound speed
small scales, characterized by their free-streaming and, of the fluid c2s . To investigate the impact of of DDE
ultimately, by the sum ofP their masses. One of the cur- models on the sum of neutrino mass, we will use two
rent tightest constraints, mν < 0.07 eV at 95% confi- functional forms for the EoS parameter:
dence level (C.L), comes from the combination of CMB
and DESI BAO data within the ΛCDM framework, show- • The Jassal-Bagla-Padmanabhan parameteriza-
ing a clear preference for the NO P [24, 30] (see also [31]). tion [14]
More recently, an upper limit of mν < 0.042 eV was
obtained using CMB data combined with DESI BAO, w(a) = w0 + wa a(1 − a) (4)
the galaxy cluster angular diameter distance, and an H0
prior from SH0ES [32]. The parameterization above consists of a combination of
These constraints are significantly relaxed within a linear and quadratic terms in the scale factor. In the
DDE framework
P modeled by the CPL parameterization, current epoch, the term −wa a2 becomes comparable to
yielding mν < 0.15 eV at 95% C.L [24]. Therefore, wa a, leading to differences at low-redshifts compared to
the current cosmological limits for the ΛCDM are on the the CPL model.
verge of excluding both inverted and normal ordering,
thereby generating tension with the results from particle • The Barboza-Alcaniz parameterization [15]
physics experiments [32, 33]. This work examines how
1−a
other well-established EoS parameterizations, such as the w(a) = w0 + wa (5)
Barboza-Alcaniz (BA) and Jassal-Bagla-Padmanabhan a2 + (1 − a)2
P
(JBP) parameterizations, impact the mν . In our anal-
ysis, we perform a Markov Chain Monte Carlo (MCMC) This parameterization exhibits a linear behavior at low-
analysis combining the latest CMB, SN, and DESI BAO redshifts. Unlike the CPL parameterization, which blows
data. The paper is organized as follows. In Section 2, we up exponentially as the scale factor approaches infinity
detail the functional forms of the DDE models consid- (for wa > 0), the BA parameterization remains well-
ered. In Section 3, we present the methodology adopted behaved at all times, while still allowing for deviations
and the results obtained. In Section 4, we present and from the CPL scenario2 .
discuss our main conclusions.
III. ANALYSIS AND RESULTS

II. DYNAMICAL DARK ENERGY MODELS

To obtain constraints on the cosmological parame-
ters for the DDE models above, we perform a Monte
DDE models affect the evolution of the Universe in two Carlo Markov Chain (MCMC) numerical analysis using
primary ways: through the expansion rate and through the Planck 2018 temperature and polarization data (in-
the cosmological perturbations that give rise to struc- cluding low and high-multipole modes and lensing) [34],
tures. In the case of the Universe’s expansion rate, as- along with SN data from Pantheon+ [35, 36] and geomet-
suming a statistically homogeneous and isotropic Uni- ric BAO information from the DESI collaboration [24].
verse that follows the Friedmann-Lemaı̂tre-Robertson- For this, we implement the two parameterizations in the
Walker (FLRW) metric, we have: Boltzmann solver code Cosmic Linear Anisotropy Solv-
ing System (CLASS) [37], in which we interfaced with
H 2 = H02 ρr a−4 + ρm a−3 + ρde a−3 exp [Υ(a)] ,
 
(1)

where w(a) represents

hR the EoS as a ifunction of the scale
a 2 Among other well-known EoS parameterizations, the BA gives
factor and Υ(a) = 1 −3 w(a) a da . For the cosmolog- the most significant improvement over the ΛCDM model when
ical perturbations the first and most obvious impact of a analyzed in the context of current CMB, BAO, and SN observa-
time-evolving EoS is on the density perturbation and on tions [25].
 3

Figure 2. w0 −wa plane for the JBP, BA and CPL parameter-

izations with Planck 2018 temperature, polarization and lens-
ing together with Pantheon+ SN and DESI BAO data. The
dashed line intersection represents the SCM values, wa = 0
and w0 = −1. The filled contours are for the models assuming
the neutrino mass as a free parameter (Blue, Red and Green),
while the black empty contours (shown here for completeness)
Figure 1. Confidence contours at 68% and 95% C.L. for the stand for the JBP model with a fixed neutrino mass.
JBP and BA parameterizations with Planck 2018 tempera-
ture, polarization and lensing together with Pantheon+ SN
and DESI BAO data. neutrino mass as a free parameter, the JBP model pro-
gressively aligns with the ΛCDM scenario, in contrast
to the BA model, which remains more than 2σ away.
the Cobaya code to perform the MCMC statistical anal- Similarly, the CPL model shows that allowing for free
ysis [38, 39]. neutrinos has a slight influence on its analysis, consistent
Figure 1 shows the 68% and 95% confidence level con- with the results reported by the DESI collaboration. This
tours obtained from our analysis. It is immediately ev- result suggests that relaxing constraints on the dark en-
ident that the main differences between them lie in the ergy parameters may help constrain the neutrino mass,
EoS parameters. It shows P significant differences, par- although the correlation is not particularly strong.
ticularly in w0 , wa , and mν . Table I presents each Figure 3 shows the EoS evolution with redshift for
model’s mean values and associated errors. The 2σ limits each model, using Planck CMB temperature, polariza-
are provided for the sum of the neutrino masses, while tion, and lensing data, along with Pantheon+ SN and
the 1σ limits are shown for the other parameters. For DESI BAO. They behave similarly at low redshifts, re-
the
P BA model, we find an upper 95% one-tail limit of maining in the quintessence region, i.e., w(z) > −1. At
mν < 0.18 eV. The JBPPmodel provides a more re- high redshifts, there are significant differences. The BA
strictive upper limit, with mν < 0.13 eV. Using the model crosses at a slightly lower redshift and remains in
same data set,Pthe CPL parameterization yields an up- the phantom region near w(z) = −1. Interestingly, for
per bound of Pmν < 0.18 eV 3 . These results demon- the JBP model, it tends to cross w(z) = −1 twice. It
strate that the mν constraints depend on the param- leaves the quintessence region and remains very close to
eterization adopted in the analysis. However, for those w(z) = −1 with a trend toward the phantom region until
considered in this study, the limits obtained alleviate the it returns to the quintessential region for high z values.
tension and agree with the mass hierarchy scheme4 . These EoS behaviors for these models are similar to those
In Figure 2, we show the w0 -wa plane. When treating presented in the work [25], where a similar analysis was
performed with the neutrino mass fixed. The most sig-
nificant difference is in the JBP behavior, which shows a
3
larger error in w(z).
Our neutrino mass constraints for the CPL model are slightly
different from the one reported in the DESI paper [24]. Here, we
use CMB data from the Planck baseline likelihood only, while
the DESI collaboration uses Planck combined with Data Release IV. FINAL REMARKS
6 of the Atacama Cosmology Telescope.
4 For completness, we also conduct analyses using a logarithmic
parameterization defined as w(a) = w0 + wa ln(a). Although not Neutrinos are one of the most promising probes to un-
presented here, the results derived from this approach are similar veil aspects of the physics at high energies. For instance,
to those obtained by the BA parameterization. in the well-known seesaw paradigm [40–43], the magni-
 4

Parameter BA JBP CPL

h 0.6808 ± 0.0073 0.6792 ± 0.0071 0.6801 ± 0.0071
Ωm 0.3080 ± 0.0071 0.3082 ± 0.0069 0.3085 ± 0.0069
Ωb 0.0484 ± 0.0011 0.0486 ± 0.0010 0.0484 ± 0.0010
ns 0.9657 ± 0.0040 0.9667 ± 0.0038 0.9658 ± 0.0037
σ8 0.816 ± 0.012 0.814 ± 0.011 0.815+0.012
−0.011
0.072+0.11 0.050+0.083 0.070+0.11
P
mν (eV) −0.074 (2σ) −0.051 (2σ) −0.073 (2σ)
w0 −0.858 ± 0.055 −0.795 ± 0.089 −0.813 ± 0.064
wa −0.36+0.17
−0.13 −1.30 ± 0.60 −0.73+0.34
−0.26

Table I. Constraints at 2σ level on the sum of the neutrino masses and 1σ level for all other parameters obtained with Planck
2018, Pantheon+ and DESI for the BA, JBP and CPL parameterizations.

Figure 3. EoS evolution obtained for Planck+DESI+Pantheon+ data assuming the BA (Blue) and JBP parameterizations
(red). The dashed line indicates the ΛCDM model, w(z) = −1.

tude of the neutrino masses is dictated by a new en- framework.

ergy scale, associated with lepton number violation. The In this work, we investigated the robustness of the mν
mechanisms that lead to this symmetry breaking pro- tension and the results obtained from the CPL model.
cess may be associated with Grand Unification theories By performing a MCMC analysis with the most recent
[44], and also may offer a solution to the fermion/anti- available BAO data from DESI along with temperature,
fermion unbalance of the universe [45]. Meanwhile, the polarization, and lensing data from Planck and Type
micro-physical aspects of the standard (active) and sterile Ia SN from Pantheon+, we studied how different well-
neutrinos are set by the Yukawa structure of the model’s known EoS parameterizations, such as BA and P JBP pa-
Lagrangian, which may be constrained by the oscillation rameterizations, influence the estimates of mν . We
process observed in the atmospheric and solar neutrinos found that the BA and JBP follow the same trend as
and also set the possible signatures left by new particles the CPL by helping to relax the constraint on the sum
in related experiments [46, 47]. of neutrino masses. The P neutrino mass upper limit ob-
In this regard, the complementarity of cosmological tained for the BA is mν < 0.18 eV while the JBP
and terrestrial observations is an important aspect for parameterization
P provides a more restrictive upper limit
constraining the theory. While neither type of exper- of mν < 0.13 eV, a difference of ∼ 30%. We also
iment can determine their fundamental properties in- note that the main difference lies primarily in the con-
dependently, their combination can help narrow down straints on the EoS parameters. This result suggests that
the possible mass scale and provide insights into their a larger uncertainty on w0 and wa values leads to tighter
interactions. Thus, as the tension between neutrino constraints on the neutrino mass. For the JBP param-
mass measurements derived from cosmological data in eterization, we obtain constraints on w0 and wa that
the ΛCDM framework and those obtained from precise are compatible with the ΛCDM model values within 2σ.
ground-based experiments becomes increasingly appar- However, this is not the case for the BA, which remains
ent, it is crucial to explore alternative models to ad- more than 2σ away.
dress this inconsistency. In this context, the DDE sce- Finally, our results
P confirm the ongoing tension in the
nario modeled by the CPL parameterization has demon- measurements of mν within the context of the stan-
strated
P the potential to resolve this discrepancy, yield- dard cosmological model, as reported by [32, 33]. They
ing mν < 0.15 eV at a 95% (CL) [24]. Additionally, also demonstrate that dynamical dark energy models can
parametric analyses of recent data from DESI indicate a effectively alleviate this issue. Furthermore, our analysis
P
slight preference for this model compared to the ΛCDM indicates that the degree to which constraints on mν
 5

are relaxed depends moderately on the chosen parame- ACKNOWLEDGEMENTS

terization.
GR is supported by the Coordenação de Aper-
feiçoamento de Pessoal de Nı́vel Superior (CAPES). JR
thanks the Fundação de Amparo à Pesquisa do Estado do
Rio de Janeiro (FAPERJ) grant No. E-26/200.513/2025.
JSA is supported by CNPq grant No. 307683/2022-2 and
FAPERJ grant No. 259610 (2021). We also acknowledge
the use of the Cobaya, CLASS and GetDist. This work
was developed thanks to the computational support of
the National Observatory Data Center (CPDON).

[1] S. Weinberg, Rev. Mod. Phys. 61, 1 (1989). 013–013.

[2] S. Weinberg, in 4th Int. Symposium on Sources and De- [22] W. Yang, S. Pan, E. Di Valentino, E. N. Saridakis, and
tection of Dark Matter in the Universe (2000) pp. 18–26, S. Chakraborty, Physical Review D 99, 10.1103/phys-
arXiv:astro-ph/0005265. revd.99.043543 (2019).
[3] T. Padmanabhan, Phys. Rept. 380, 235 (2003), [23] W. Yang, E. Di Valentino, S. Pan, Y. Wu, and J. Lu,
arXiv:hep-th/0212290. Monthly Notices of the Royal Astronomical Society 501,
[4] J. S. Alcaniz, Braz. J. Phys. 36, 1109 (2006), arXiv:astro- 5845–5858 (2020).
ph/0608631. [24] DESI Collaboration, Journal of Cosmology and As-
[5] E. J. Copeland, A. R. Liddle, and D. Wands, Phys. Rev. troparticle Physics arXiv:2404.03002 [astro-ph.CO].
D 57, 4686 (1998), arXiv:gr-qc/9711068. [25] W. Giarè, M. Najafi, S. Pan, E. Di Valentino, and
[6] L. Amendola, Phys. Rev. D 62, 043511 (2000), J. T. Firouzjaee, Journal of Cosmology and Astroparticle
arXiv:astro-ph/9908023. Physics 2024 (10), 035.
[7] J. A. S. Lima and J. S. Alcaniz, Mon. Not. Roy. Astron. [26] B. R. Dinda and R. Maartens, JCAP 01, 120,
Soc. 317, 893 (2000), arXiv:astro-ph/0005441. arXiv:2407.17252 [astro-ph.CO].
[8] A. Dev, D. Jain, and J. S. Alcaniz, Phys. Rev. D 67, [27] A. Sousa-Neto, C. Bengaly, J. E. González, and J. Al-
023515 (2003), arXiv:astro-ph/0209379. caniz, No evidence for dynamical dark energy from DESI
[9] J. S. Bagla, H. K. Jassal, and T. Padmanabhan, Phys. and SN data: a symbolic regression analysis (2025),
Rev. D 67, 063504 (2003), arXiv:astro-ph/0212198. arXiv:2502.10506 [astro-ph.CO].
[10] V. Sahni and Y. Shtanov, JCAP 11, 014, arXiv:astro- [28] E. Ó. Colgáin, M. G. Dainotti, S. Capozziello, S. Pouro-
ph/0202346. jaghi, M. M. Sheikh-Jabbari, and D. Stojkovic, arXiv
[11] F. C. Carvalho, J. S. Alcaniz, J. A. S. Lima, and e-prints , arXiv:2404.08633 (2024), arXiv:2404.08633
R. Silva, Phys. Rev. Lett. 97, 081301 (2006), arXiv:astro- [astro-ph.CO].
ph/0608439. [29] S. Navas et al. (Particle Data Group), Phys. Rev. D 110,
[12] M. Chevallier and D. Polarski, International Journal of 030001 (2024).
Modern Physics D 10, 213 (2001), arXiv:gr-qc/0009008 [30] R. Jimenez, C. Pena-Garay, K. Short, F. Simpson, and
[gr-qc]. L. Verde, JCAP 09, 006, arXiv:2203.14247 [hep-ph].
[13] E. V. Linder, Phys. Rev. Lett. 90, 091301 (2003), [31] S. Gariazzo et al., JCAP 10, 010, arXiv:2205.02195 [hep-
arXiv:astro-ph/0208512 [astro-ph]. ph].
[14] H. K. Jassal, J. S. Bagla, and T. Padmanabhan, Physical [32] J.-Q. Jiang et al., Journal of Cosmology and Astroparti-
Review D 72, 103503 (2005). cle Physics 2025 (01), 153.
[15] E. Barboza and J. Alcaniz, Physics Letters B 666, 415 [33] S. Gariazzo, O. Mena, and T. Schwetz, Physics of the
(2008). Dark Universe 40, 101226 (2023).
[16] E. M. Barboza, J. S. Alcaniz, Z. H. Zhu, and R. Silva, [34] Planck Collaboration, Astronomy and Astrophysics 641,
Phys. Rev. D 80, 043521 (2009), arXiv:0905.4052 [astro- A6 (2020), arXiv:1807.06209 [astro-ph.CO].
ph.CO]. [35] D. Brout et al., The Astrophysical Journal 938, 110
[17] D. Wang, Physical Review D 106, 10.1103/phys- (2022).
revd.106.063515 (2022). [36] D. Scolnic et al., The Astrophysical Journal 938, 113
[18] J. S. Peracaula, A. Gomez-Valent, and J. de Cruz Perez, (2022).
Signs of dynamical dark energy in current observations [37] J. Lesgourgues, The Cosmic Linear Anisotropy Solving
(2019), arXiv:1811.03505 [astro-ph.CO]. System (CLASS) I: Overview (2011), arXiv:1104.2932
[19] S. Pan, W. Yang, E. Di Valentino, E. N. Saridakis, and [astro-ph.IM].
S. Chakraborty, Physical Review D 100, 10.1103/phys- [38] D. Blas, J. Lesgourgues, and T. Tram, JCAP 07, 034,
revd.100.103520 (2019). arXiv:1104.2933 [astro-ph.CO].
[20] C. Escamilla-Rivera and A. Nájera, Journal of Cosmology [39] J. Torrado and A. Lewis, Journal of Cosmology and As-
and Astroparticle Physics 2022 (03), 060. troparticle Physics 2021 (05), 057.
[21] E. D. Valentino, A. Melchiorri, and J. Silk, Journal [40] P. Minkowski, Phys. Lett. 67B, 421 (1977).
of Cosmology and Astroparticle Physics 2020 (01),
 6

[41] T. Yanagida, Proceedings: Workshop on the Unified The- [44] H. Fritzsch and P. Minkowski, Annals Phys. 93, 193
ories and the Baryon Number in the Universe: Tsukuba, (1975).
Japan, February 13-14, 1979, Conf. Proc. C7902131, 95 [45] D. Bodeker and W. Buchmuller, Rev. Mod. Phys. 93,
(1979). 035004 (2021), arXiv:2009.07294 [hep-ph].
[42] M. Gell-Mann, P. Ramond, and R. Slansky, Supergrav- [46] G. Rodrigues, J. G. Rodrigues, F. B. M. d. Santos, and
ity Workshop Stony Brook, New York, September 27-28, J. S. Alcaniz, JCAP 11, 055, arXiv:2409.13918 [astro-
1979, Conf. Proc. C790927, 315 (1979), arXiv:1306.4669 ph.CO].
[hep-th]. [47] A. G. Dias, C. A. de S. Pires, P. S. Rodrigues da
[43] R. N. Mohapatra and G. Senjanovic, Phys. Rev. Lett. Silva, and A. Sampieri, Phys. Rev. D 86, 035007 (2012),
44, 912 (1980), [,231(1979)]. arXiv:1206.2590 [hep-ph].