Interpreting Dark Energy Data Away from $\Lambda$
Extracted source text · arXiv 2410.10981
\documentclass[prd, twocolumn, nofootinbib, floatfix]{revtex4-2}
\usepackage{amsmath}
\usepackage{graphicx}
\usepackage{dcolumn}
\usepackage{bm}
\usepackage{epsfig}
\usepackage{amssymb,latexsym,mathrsfs,empheq}
\usepackage{graphicx}
\usepackage{color}
\usepackage{xcolor}
\usepackage{mathtools}
\usepackage{hyperref}
\usepackage{cleveref}
\hypersetup{
colorlinks=true,
linkcolor=red,
citecolor=blue,
}
\newcommand{\be}{\begin{equation}}
\newcommand{\ee}{\end{equation}}
\newcommand{\bs}{\begin{split}}
\newcommand{\bea}{\begin{eqnarray}}
\newcommand{\eea}{\end{eqnarray}}
\newcommand{\al}{\alpha}
\newcommand{\dl}{\delta}
\newcommand{\gam}{\gamma}
\newcommand{\eps}{\epsilon}
\newcommand{\Ok}{\Omega_k}
\newcommand{\Om}{\Omega_m}
\newcommand{\om}{\Omega_m}
\newcommand{\wde}{w_{\rm de}}
\newcommand{\rde}{\rho_{\rm de}}
\newcommand{\lcdm}{$\Lambda$CDM}
\newcommand{\gm}{G_{\rm matter}}
\newcommand{\gl}{G_{\rm light}}
\newcommand{\eric}[1]{\textcolor{blue}{[{\bf EL}: #1]}}
\begin{document}
\title{Interpreting Dark Energy Data Away from $\Lambda$}
\author{Eric V.\ Linder}
\affiliation{
Berkeley Center for Cosmological Physics \& Berkeley Lab,
University of California, Berkeley, CA 94720, USA
}
\begin{abstract}
Dark energy away from a cosmological constant $\Lambda$ -- like
early universe inflation that ends -- can be understood in terms
of well defined physical behaviors. These guide dark energy into
thawing or freezing classes, with $w_0$--$w_a$ arising as a
physical calibration of the phase space. Other regions of phase space
-- zones of avoidance -- require violation of some basic principle.
We explore these cases, drawing a direct analogy with how nonGaussianity
in inflation can add physics beyond standard dynamics.
We examine the physics implications if the
best fit of current data is taken to be truth, outlining four properties, and investigate
the reality
of phantom crossing $w=-1$, finding it significantly favored.
\end{abstract}
\date{\today}
\maketitle
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\section{Introduction}
Cosmic acceleration is amazing. It bends the cosmic expansion
against the attraction of gravity, stretching distances, and
suppresses growth of large scale structure, again against the
attraction of gravity, possibly shutting growth off in the future. We
do not know its source other than that it lies outside the
Standard Model of particle physics.
The ``simplest'' explanation might be a cosmological constant $\Lambda$
but this comes with its own mysteries and paradoxes, such as why
it has neither Planck scale energy density nor zero but rather of such
a magnitude that it only became significant within the last efold
of expansion history -- moreover despite quantum corrections that
should enter at higher energy scales.
Going beyond $\Lambda$ gives dark energy some dynamics, and this
carries with it clear physical characteristics due to the dark
energy field existing in a universe with many efolds of radiation and matter
dominated expansion. Thus it is very comprehensible. (For expansion
behavior, even an origin such as a modification of gravity can be
treated adequately as an effective scalar field with potential and
kinetic terms.)
In \Cref{sec:dyn} we briefly review dark energy dynamics and
its resulting physical characteristics. \Cref{sec:mirage} considers
current data at face value and its implications of going beyond
standard principles, with \Cref{sec:nonKG} illustrating how this
tests the framework itself, in close analogy with inflation.
We look to
the future in \Cref{sec:fut} for what other types of data can
say about the physical origin of dark energy.
%%%%%%%%%%%%%%%%%%
\section{Physics of Dark Energy Dynamics} \label{sec:dyn}
Dark energy does not exist in a vacuum. Less waggishly, we
understand the basics of dark energy behavior very well because
it has evolved not alone but in an environment of many efolds of
radiation and matter dominated expansion. This forces it to have
certain dynamics that cannot be avoided without fine tuning (or
violating some standard principles, as discussed in the next
section).
The immense Hubble friction from the environment governs the
dark energy dynamics, together with steepness of the dark energy
potential. Put most simply (see \cite{caldlin,paths,flow} for
details) if the Hubble friction dominates then the (effective)
field is frozen in place, released to thaw only when the universe
expands sufficiently to dilute radiation and matter and reduce
the Hubble friction. Such thawing fields evolve away from cosmological
constant behavior. If instead the potential has sufficiently steep slope, the field rolls in the early universe until eventually
it approaches a region of shallow slope near the potential minimum.
The Hubble friction takes over and the field slows. Such freezing
fields evolve toward cosmological constant behavior.
One could plot these behaviors in field phase space $\phi$--$\dot\phi$
but it is more convenient to work with the dark energy pressure to
energy density, or equation of state, ratio $w$ and its efold derivative
$w'$. This is closer to the Hubble parameter $H=\dot a/a$ describing
cosmic expansion history since
\be
\left(\ln H^2\right)'\equiv\frac{d\ln H^2}{d\ln a}=
-3\left[1+\sum w_i(a)\,\Omega_i(a)\right]\ .
\ee
Furthermore, it is cleaner if the dark energy is not actually
a scalar field.
\Cref{fig:wwp} illustrates how these physical effects of Hubble friction
and potential steepness, operating over many efolds of radiation and
matter domination, bring the dark energy evolution into narrow regions
of the phase space, i.e.\ the thawing and freezing classes.
%%%%%%%%%%%%%%%%%%%%%%
\begin{figure}
\centering
\includegraphics[width=\columnwidth]{wwpLabeled.jpg}
\caption{Simple physics in an expanding universe brings
dark energy dynamics into narrow regions of the phase
space, either thawing or freezing. Adapted
from \cite{calde}.}
\label{fig:wwp}
\end{figure}
What about the rest of phase space? In order for dark energy at
present to lie in another region, some standard physical principle
must be broken. \Cref{fig:zone} defines the ``zones of avoidance''.
The middle region, between the thawing and freezing regions, requires
fine tuning, where the two physical effects must -- coincidentally
just at present -- nearly balance. The high region violates early
radiation and matter domination, i.e.\ the many efolds of Hubble
friction. The low region requires the field to evolve noncanonically,
e.g.\ roll upslope. On the left half of the diagram the field is in
the phantom regime, $w<-1$, e.g.\ with a negative kinetic term.
We will revisit the phantom and high regions in the next section.
%%%%%%%%%%%%%%%%%%%%%%
\begin{figure*}
\centering
\includegraphics[width=\textwidth]{wwpZones.jpg}
\caption{Zones of avoidance, outside the thawing and
freezing regions, can only be accessed by violating some
basic physical principle. The text describes these for
each region.}
\label{fig:zone}
\end{figure*}
Thus we know quite a lot about dark energy characteristics even
without identifying dark energy in specific. The $w$--$w'$ phase
space provides an excellent guide to the physics affecting dark
energy. However, cosmological data cannot generally fit the two
full functions $w(a)$, $w'(a)$. Indeed, observations depend on
(multiple) integrals over $w(a)$ so the data does not see the details
of the functions. Principal component analysis, or similar techniques,
demonstrate that only two numbers, not full free functions, can be
fit \cite{pca}, at least until observations improve beyond 0.1\%
precision.
The art, then, is choosing the right two quantities to preserve
the essential physics information. For our parameter corresponding
to $w(a)$, let us choose the value today, $w_0\equiv w(a=1)$. To
describe $w'(a)$, let's try stretching the time axis, i.e.\ scaling
$w'(a)$. \Cref{fig:wa} shows the results -- a remarkable success!
Different varieties of theories are tightly calibrated, and in
particular the wide variety of different models in the thawing class
come together to form a universal behavior in this new phase space.
(See \cite{calde} for more details.)
%%%%%%%%%%%%%%%%%%%%%%
\begin{figure*}
\centering
\includegraphics[width=\textwidth]{wacal.jpg}
\caption{Dark energy evolution can be further calibrated
from the thawing/freezing wedges in $w$--$w'$ of
\Cref{fig:wwp} [see left panel] into tight, near universal
behavior through stretching the time coordinate [see right
panel]. This physical calibration is the foundation of
$w_0$--$w_a$. (Note the sign flip of $w_a$
relative to $w'$, interchanging the top and
bottom half planes.) Adapted from \cite{calde}.}
\label{fig:wa}
\end{figure*}
This is the meaning of $w_0$--$w_a$, with $w(a)=w_0+w_a(1-a)$.
Created as a fit to exact
solutions of scalar field dynamics in \cite{linexp} -- never as
an arbitrary parametrization or ``Taylor expansion'' -- it then
was demonstrated to be a calibration form, a physics based
encapsulation of the full $w$--$w'$ phase space in \cite{calde}.
%%%%%%%%%%%%%%%%%%
\section{Interpreting Dark Energy Data} \label{sec:mirage}
Now that we have the tools in place: that $w_0$--$w_a$ faithfully
describes dark energy physics, what standard principles predict in
terms of the expected regions of thawing and freezing, and what
violations must occur to avoid those regions, let us consider the
current data.
The Dark Energy Spectroscopic Instrument (DESI \cite{desi}) recently
released cosmological analyses \cite{kp7,crossing,focus} of its
highly precise first year baryon acoustic oscillation (BAO) distance
measurements, supplemented with supernova (SN) distances and cosmic
microwave background (CMB) data. The analyses -- starting with blinded
BAO data -- were done in several different ways and all came to a
similar conclusion: $\Lambda$CDM is disfavored at $\sim2.5$--3.9$\sigma$
and the best fit lies neither in the thawing nor freezing regions, but
what we here call the high region.
The core principle violated in the high region is that of many efolds
of dark energy evolution in a radiation and matter dominated universe.
This is quite astounding, as that is the principle we might consider
least likely to be broken (as opposed to, say, the scalar field
kinetic structure), since it is indicated by a vast array of
cosmological observations. One possibility is that there are
statistical fluctuations or systematic offsets in the data, but
this is somewhat alleviated by the blind nature of the analysis
and the robustness of the result under different analysis approaches.
Therefore we will focus here on the implications if the
dark energy properties are truly as indicated by the best fit values,
say $w_0\approx-0.64$, $w_a\approx-1.27$. (We will not worry here
about specific values, only about what region of the phase space.
And we will keep in mind the ``if'': recall that $3\sigma$ is not
definitive, and the DESI analysis mindfully used the phrasing ``tantalizing
hint of deviations'' from $\Lambda$.)
Taking the best fit as truth requires the dark energy to possess
four properties:
\begin{itemize}
\item The dark energy was phantom at $z\gtrsim1$;
\item The dark energy must ``superevolve'' faster than the Hubble
friction from matter domination would indicate;
\item The dark energy crossed $w=-1$, ``the phantom divide'';
\item The dark energy recently evolved toward $w>-1$, i.e.\ away from
$\Lambda$.
\end{itemize}
We know the physics to give dark energy each individual property,
but they generally don't all go together!
Beyond all these aspects concerning the general region of phase space,
the best fit lies along a very special cut through phase space
call mirage dark energy. Mirage dark energy is a continuous
family of dark energy
that extends from $\Lambda$ into the high region, along the
line $w_a\approx -3.66(1+w_0)$ \cite{mirage}. It has several interesting
properties, including that the distance to CMB last scattering is
nearly identical for all members of the family (and so if one
naively fit only a constant $w$, rather than looking at the
dynamics, one would find $\langle w\rangle\approx-1$).
In seeking the physics behind the dark energy, an important
question is whether the apparent interpretation regarding
the phantom crossing is robust. That is, does $w(a)$ really
cross $-1$ or does it just appear that way because of the values
of $w_0$--$w_a$? This has been partly addressed in
\cite{crossing} by using a nonparametric $w(a)$, and in
\cite{focus} by using a $w(a)$ that does not cross $-1$, showing that the crossing is real,
but here we revisit it more quantitatively.
We know that values of $w_0$--$w_a$ characteristic of the
thawing region (not the high region) can be treated by the
algebraic thawing form \cite{algthaw} with $w(a)>-1$ always, with
the two approaches agreeing to ${\mathcal O}(0.1\%)$.
However, this does not
work for the strong mirage case where $|w_a|\gtrsim 1$.
Using the algebraic thawing form with $p=5$ (rather than the
$p=1$ for thawing dark energy, in order to give a sharp
evolution and hence large $|w_a|$) yields a $w(a)>-1$ at all
times and yet $w_0$--$w_a$ parameters that approximate our
best fit values. (See \cite{steinh} for a similar approach.)
Fitting all the data to either the thawing with $w_0$--$w_a$
or the algebraic thawing forms gives no discernible
difference. Their $\Delta\chi^2<0.3$ (also see
\cite{focus}). However, fitting all
the data to the mirage or the algebraic form with
$p=5$ yields a clear distinction. While the mirage case
gives a superb fit (since the best fit to the data indeed
lies nearly on the mirage line), the algebraic form --
where phantom crossing is merely a spurious interpretation
due to $w_0$--$w_a$ -- suffers by $\Delta\chi^2>9$,
i.e.\ $\sim3\sigma$ worse. Thus the data favors that
the phantom crossing is indeed a real property.
%%%%%%%%%%%%%%%%%%%%%%%%%%%
\section{Interpreting Dark Energy Physics} \label{sec:nonKG}
Thus the four conditions itemized above describe
the physics needed. Before pursuing the specific physics
indicated, let us look at the bigger picture. What we are
really doing is testing the framework, not fitting
parameters or models. Any cosmic expansion history can be
treated as an effective scalar field, i.e.\ $w$--$w'$. So
data giving a result in one of the Zones of Avoidance is
testing the physics of the scalar field, not just one
particular model. In gist, we are checking the dynamical
(Klein-Gordon) equation itself.
Note that this is very similar to what is done with early
universe inflation. Beyond the dynamics (e.g.\ the scalar
spectral tilt $n_s$ and primordial gravitational wave or
tensor to scalar ratio $r$), we look at for example
inflationary nonGaussianity. Here we are looking at dark
energy nonKlein-Gordonness.
As in inflation, this can arise through any of the
following physical characteristics:
\begin{itemize}
\item Noncanonical kinetic term
\item Multiple, interacting fields
\item Nonstandard vacuum
\end{itemize}
A noncanonical kinetic term seems unlikely to be the
(sole) cause of data preferring the high region in
phase space, because such k-essence lives in the low
region (or its mirror in the left half plane). Multiple,
interacting fields is a possibility in fact richer than
its inflation analog, since dark energy could interact
not only with another scalar field but matter fields
(which are negligible in the inflationary phase).
While for inflation
a nonstandard vacuum means not having an initial
Bunch-Davies state, for dark energy it refers to the
scalar field either having a negative potential or
arising through a phase transition.
To achieve all four physical conditions needed from
the previous section, we can now see a possibility:
dark energy that originates in a phase transition
(e.g.\ vacuum metamorphosis \cite{vm1,vm2,vm3}) at
redshift $z\approx1-2$
(thus avoiding the ``many efolds of dark energy evolution in a radiation and
matter dominated universe'' -- not because
of violating matter domination but due to
hiding from the many efolds),
so it superevolves, rapidly
increasing its energy density from near zero (so
$w_a\lesssim-1$) and heading from deeply phantom
toward $w=-1$. But we need something more: most
phase transition models asymptote to a de Sitter
future, with $w=-1$ (in the vacuum metamorphosis
case this is because the Ricci scalar goes to a
constant, locking into the value given by the
scalaron mass). And indeed such ``emergent'' dark
energy models do not fit the data as well as
mirage dark energy (which crosses $w=-1$), e.g.\
\cite{focus}.
The second element needed is some coupling that
brings the field across $w=-1$ so that today
$w_0>-1$. Indeed, particle physics generally gives
interactions unless some symmetry or other principle
suppresses it. Thus one possible physics
framework for the
desired mirage dark energy is a phase transition
with the resulting field having a natural coupling.
Such a theory would promise a rich array of tests
from observational data.
%%%%%%%%%%%%%%%%%%
\section{More Data for More Physics} \label {sec:fut}
More data will not only test whether the best fit
maintains its position in the high region of phase
space, but can offer different angles on the physics.
Just as in inflation one seeks to measure primordial
gravitational waves and look for consistency between
$r$ and $n_s$, for dark energy one can contrast growth
probes with distance probes.
More generally one can explore the gravitational sector,
in terms of
\begin{itemize}
\item Growth vs expansion
\item Gravitational effects on matter and light
\item Gravitational friction on gravitational wave
propagation
\end{itemize}
The second and third investigate gravity beyond general
relativity, with the second looking for time and space
dependence in the strength of gravitational coupling,
i.e.\ $G_N\to G_{\rm matter}(k,a)$, $G_{\rm light}(k,a)$,
and the third checking for a time variation of the
Planck mass through a gravity friction term added to
the Hubble expansion friction in gravitational wave
propagation, causing the wave amplitude to decay
differently than the standard 1/distance.
Thus, measurements of growth of large scale structure,
gravitational lensing, and gravitational wave distances
from standard sirens all enable new tests of the
physics framework. Note that the best fit we have been
considering does not require any modification to general
relativity though, and there may be no signal in the
second and third items.
General relativity does give a fixed relation between
the growth factor of large scale structure and the
cosmic expansion history. So growth observations such
as from redshift space distortions and peculiar velocities
can provide a test of the predicted consistency, whether
the cosmology is that of the best fit, or of $\Lambda$.
That is important, but if the growth data shows consistency
with the best fit it does not necessarily zero in on
the physical origin.
Growth data does tend to constrain dark energy with a
different degeneracy direction, i.e.\ weighting of $w_0$
vs $w_a$ say, than distance data. One can partly view
this as that distances are measured from the observer
back in history to the source, while growth depends on
evolution from the early universe to the tracer.
Thus the constraints on dark energy properties are
often tighter when combining growth and distance data.
However, there is one notable exception! Mirage dark
energy -- exactly what the best fit seems to indicate --
is mostly exempt from this. Its particular properties
give that the growth leverage direction essentially
overlaps with that of the distance to CMB last scattering,
and no major gain is expected. This is illustrated in
\Cref{fig:grocmb}. So sadly, growth data will not have as
much of an impact on cosmological constraints as it
would in a universe that happened to be off the
mirage line.
Nevertheless, new data -- of expansion, growth, and
gravity -- will be highly useful for crosschecking
the current ``tantalizing hints'', testing general
relativity, and gradually narrowing in on parameter
values.
%%%%%%%%%%%%%%%%%%%%%%
\begin{figure}
\centering
\includegraphics[width=\columnwidth]{grocmb.jpg}
\caption{Matching the CMB distance to last scattering is
closely related to matching the growth history, in
particular the linear growth factor ratio
$g(a = 0.35)/g(a = 1)$. Thus for a universe involving
mirage dark energy behavior, growth data such as from
redshift space distortions would add only modest leverage
to distance data. Adapted from \cite{mirage}.}
\label{fig:grocmb}
\end{figure}
%%%%%%%%%%%%%%%%%%%%%%%%
\section{Conclusions}
Dark energy characteristics can be well understood as the
result of physics during evolution through a long period
of radiation and matter domination. The impact of dark
energy on observations is superbly accounted for by
means of the physical calibration of the phase space via
$w_0$--$w_a$.
We also understand the ``zones of avoidance'' in dark
energy phase space and what standard principles would need
to be violated to allow those regions to come into play.
Amazingly, current data gives tantalizing hints that indeed
dark energy behaves as if it lies in the high region zone
of avoidance. Dark energy is best fit for current data by mirage dark energy,
which has several unusual properties including phantom
crossing and a non-de Sitter future state, and moreover
exhibits a rapid evolution violating having a long history of
evolution during matter domination. A possible physical
scenario is identified of a phase transition, with the
resulting field coupled in some manner to another (possibly
matter) field. We emphasize that the same physical
characteristics have been found in approaches not utilizing
$w_0$--$w_a$ but more model agnostic forms. Furthermore
we can verify that the phantom crossing being a real, not
merely apparent, characteristic is preferred by the data
by $\gtrsim3\sigma$.
In analogy to nonGaussianity in inflation we can identify
physical origins for violating the standard Klein-Gordon
evolution. Nonstandard vacuum in the form of
a phase transition is a contender. Similarly we can mirror inflationary consistency
tests to reveal more about the nature of dark energy,
especially through gravity beyond general relativity.
We also note that due to the unique properties of mirage
dark energy, growth of structure observations will tend
to give more incremental rather than orthogonal
improvements.
Nevertheless, the data that is forthcoming of new and
higher accuracy expansion, growth, and gravity observations
will test the tantalizing hints of strong mirage dark
energy. Other methods searching for a phase transition
could yield interesting contributions. In particular,
the next generation of experiments aiming for a higher
redshift range, e.g.\ DESI2 \cite{desi2}, SpecS5
\cite{specs5}, could explore whether dark energy density
was vanishingly small at $z\gtrsim2$ \cite{rise,sailer},
indicating that dark energy arose rapidly, such as from
a phase transition.
\acknowledgments
I thank colleagues at ICTP (Trieste), IFPU (Trieste),
Roma La Sapienza (Rome), UiO (Oslo), LMU (Munich), IAA
(Granada), US (Seville), UAM (Madrid), CGWU (Seoul),
and KAS (Korea) for hospitality and discussions that
helped motivate and crystallize this work. I am grateful
to Kushal Lodha for computing the algebraic case posteriors.
\begin{thebibliography}{99}
\bibitem{caldlin}
R.R. Caldwell, E.V. Linder, The Limits of Quintessence,
Phys. Rev. Lett. 95 (2005) 141301
[arxiv:astro-ph/0505494]
\bibitem{paths}
E.V. Linder, The Paths of Quintessence,
Phys. Rev. D 73, 063010 (2006) [arXiv:astro-ph/0601052]
\bibitem{flow}
R.N. Cahn, R. de Putter, E.V. Linder,
Field Flows of Dark Energy,
JCAP 0811, 015 (2008) [arXiv:0807.1346]
\bibitem{pca}
E.V. Linder, D. Huterer, How Many Dark Energy Parameters?,
Phys. Rev. D 72, 043509 (2005) [arXiv:astro-ph/0505330]
\bibitem{calde}
R. de Putter, E.V. Linder,
Calibrating Dark Energy,
JCAP 0810, 042 (2008) [arXiv:0808.0189]
\bibitem{linexp}
E.V. Linder,
Exploring the Expansion History of the Universe,
Phys. Rev. Lett. 90, 091301 (2003) [arXiv:astro-ph/0208512]
\bibitem{desi}
\url{https://desi.lbl.gov}
\bibitem{kp7}
DESI Collaboration, DESI 2024 VI: Cosmological Constraints from
the Measurements of Baryon Acoustic Oscillations, arXiv:2404.03002
\bibitem{crossing}
R. Calderon et al., DESI 2024: Reconstructing Dark Energy using
Crossing Statistics with DESI DR1 BAO data, JCAP accepted, arXiv:2405.04216
\bibitem{focus}
K. Lodha et al., DESI 2024: Constraints on Physics-Focused
Aspects of Dark Energy using DESI DR1 BAO Data, arXiv:2405.13588
\bibitem{mirage}
E.V. Linder, The Mirage of $\Lambda$, arXiv:0708.0024
\bibitem{algthaw}
E.V. Linder, Quintessence’s last stand?, Phys. Rev. D 91 (2015) 063006 [arXiv:1501.01634]
\bibitem{steinh}
D. Shlivko, P. Steinhardt, Assessing observational constraints
on dark energy, Phys. Lett. B 855, 138826
(2024) [arXiv:2405.03933]
\bibitem{vm1}
L. Parker, A. Raval,
New Quantum Aspects of a Vacuum-Dominated Universe,
Phys. Rev. D 62, 083503 (2000) [arXiv:gr-qc/0003103]
\bibitem{vm2}
R.R. Caldwell, W. Komp, L. Parker, D.A.T. Vanzella,
A Sudden Gravitational Transition,
Phys. Rev. D 73, 023513 (2006) [arXiv:astro-ph/0507622]
\bibitem{vm3}
E. Di Valentino. E.V. Linder, A. Melchiorri,
A Vacuum Phase Transition Solves $H_0$ Tension,
Phys. Rev. D 97, 043528 (2018) [arXiv:1710.02153]
\bibitem{desi2}
D.J. Schlegel et al., A Spectroscopic Road Map for Cosmic
Frontier: DESI, DESI-II, Stage-5, arXiv:2209.03585
\bibitem{specs5}
\url{https://spec-s5.org}
\bibitem{rise}
E.V. Linder, The Rise of Dark Energy, arXiv:2106.09581
\bibitem{sailer}
N. Sailer, E. Castorina, S. Ferraro, M. White,
Cosmology at high redshift - a probe of fundamental
physics, JCAP 2112, 049 (2021) [arXiv:2106.09713]
\end{thebibliography}
\end{document}