arXiv:0712.4152v1 [astro-ph] 26 Dec 2007
IL NUOVO CIMENTO
Vol. ?, N. ?
The CODEX-ESPRESSO experiment: cosmic dynamics, funda-
mental physics, planets and much more...
S. Cristiani(1)(2), G. Avila(3), P. Bonifacio(1)(13), F. Bouchy(9), B. Carswell(5),
S. D’Odorico(3), V. D’Odorico(1), B. Delabre(3), H. Dekker(3) M. Dessauges(4),
P. Dimarcantonio(1), R. Garcia-Lopez(6), A. Grazian(8), M. Haehnelt(5),
J. M. Herreros(6), G. Israelian(6), S. Levshakov(10), J. Liske(3), C. Lovis(4),
A. Manescau(3), E. Martin(6), M. Mayor(4), D. Megevand(4), P. Molaro(1),
M. Murphy(5)(12), L. Pasquini(3), F. Pepe(4), J. Perez(6), D. Queloz(4), R. Rebolo(6),
P. Santin(1), P. Shaver(3), P. Span` o(7), S. Udry(4), E. Vanzella(1), M. Viel(1),
M. R. Zapatero(6), F. Zerbi(7), S. Zucker(11)
(1) INAF/National Institue for Astrophysics, Osservatorio di Trieste, Italy
(2) INFN/National Institute for Nuclear Physics, Trieste, Italy
(3) European Southern Observatory, Garching, Germany
(4) Observatoire de Gen` eve, Switzerland
(5) Institute of Astronomy, University of Cambridge, Cambridge, UK
(6) Instituto de Astrofisica de Canarias, Spain
(7) INAF/National Institue for Astrophysics, Osservatorio di Merate, Italia
(8) INAF/National Institue for Astrophysics, Osservatorio di Roma, Italia
(9) Institut d’Astrophysique de Paris, France
(10) Ioffe Physico-Technical Institute, St.Petersburg, Russia
(11) Dept. of Geophysics & Planetary Sciences, Tel Aviv University, Israel
(12) Centre for Astrophysics & Supercomputing, Swinburne University of Technology, Mel-
(13) GEPI, Observatoire de Paris, CNRS, Universit´ e Paris Diderot, Meudon, France
by the E-ELT, the most powerful telescope ever conceived, will for the first time
provide the possibility of directly measuring the change of the expansion rate of the
Universe with time and much more, from the variability of fundamental constants to
the search for other earths. A study for the implementation at the VLT of a precursor
of CODEX, dubbed ESPRESSO, is presently carried out by a collaboration including
ESO, IAC, INAF, IoA Cambridge and Observatoire de Gen` eve. The present talk is
focused on the cosmological aspects of the experiment.
CODEX, a high resolution, super-stable spectrograph to be fed
c ? Societ` a Italiana di Fisica
S. CRISTIANI ETC.
1. – Introduction
Hubble’s discovery of the expansion of the Universe, enabled in the late 20s by the
new observational facilities on Mount Wilson, brought to an end the cherished belief held
by most physicists of the time, including Albert Einstein, that the Universe is static and
The discovery has later been impressively confirmed by a vast range of astronomical
observations and led to the now widely accepted Hot Big Bang Theory, which rests on firm
pillars such as the detection of the relic Cosmic Microwave Background (CMB, [10, 13])
and the experimental verification of the prediction for the synthesis of light elements .
The Hot Big Bang is now an essential aspect of the cosmological standard model and
the central question has become: what is the stress-energy tensor of the Universe?, which
- with the ansatz of homogeneity and isotropy - reduces to: what is the mean density
and the equation of state of each mass-energy component of the Universe? Since these
parameters determine both the evolution with time and the geometry of the metric that
solves the Einstein equation one can use a measurement of either to infer their values.
The above question has been addressed by various observations: from CMB and SN
Type Ia to the abundance of clusters, from weak lensing to the large scale structure
traced by galaxies and the inter-galactic medium (IGM), and the results are consistent
with the so-called Concordance Model, a Friedmann-Robertson-Walker Universe with
no curvature, which has provided a first moderate embarrassment, due to a significant
component of dark matter whose nature remains elusive, and a major surprise with the
discovery that the expansion of the Universe has recently begun accelerating for physical
reasons basically unclear at the moment. The latter is accommodated by modifying
the stress-energy tensor to include a new component with negative pressure, something
reminiscent of the old ether, falsified by the experiment of Michelson and Morley.
It is important to note that the above mentioned experiments (CMB, Lensing, SNIa,
etc.) are based on geometry, some of them require a prior on spatial curvature and a
detailed understanding of the growth of density perturbations, hence a specific cosmolog-
ical model. For this reason, measuring the dynamics of the Universe in a clean a direct
way would be a new fundamental test of General Relativity and Cosmology.
2. – Probing the nature of the universal expansion
Is it possible to directly measure the history of the expansion? The goal is to recon-
struct the evolution of a(t), the scale factor as a function of time. We ordinarily measure
a(z) and to recover the unknown a(t) we need to know da/dt(z). In other words we need
to measure the Hubble constant, H(z) ≡ ˙ a/a using dynamics.
A way to implement this experiment is to monitor the redshifts of cosmological
sources. The change of these redshifts as a function of the observer’s time is a direct
signal of the acceleration of universe’s expansion and hence of its dynamics.
The variation of the redshift as a function of time [5, 4], turns out to be a simple
difference between the Hubble constant at the present epoch and the one at the time of
˙ z = (1 + z) Ho− H(te). (1)
The practical problem is the smallness of the signal. Fig. 1 shows what is expected for
a concordance-like model: we are dealing with accelerations of the order of 1 cm/s/yr!
THE CODEX-ESPRESSO EXPERIMENT: COSMIC DYNAMICS, FUNDAMENTAL PHYSICS, PLANETS AND MUCH MORE...3
Fig. 1. – Evolution of ˙ z as a function of redshift. The cosmological parameters have been fixed
to Ωtot = 1, H0 = 70 and different values of ΩΛ have been considered. The lines represent the
behavior of the Eq. 1, assuming the observation of the Lyman-α line.
The signal is small but displays a characteristic signature: a pattern of shrinking of the
spectrum at high-z and stretching at low-z. This is typical of a non-zero cosmological
constant: the zero crossing depends only on the ratio ΩΛ/Ωm(zo≃ 2 in a Concordance
Cosmology) and the amplitude on Ho.
3. – Measuring the Cosmic Signal
How to measure this signal? Masers (e.g. ) or molecular lines could appear very
good candidates: lines are sharp and wavelengths can be measured with very high ac-
curacy, due to their rather long wavelength, noise is also much less of a problem than
at optical wavelength. However, these targets turn out not to be well suited, since they
typically reside in deep potential wells and are subject to large peculiar accelerations, of
the same order or larger than the cosmic signal.
Absorptions from the many intervening lines in front of high-z QSOs are the most
promising candidates. Simulations, observations and analysis all concur in indicating
that the Lyman forest and associated metal lines are produced by systems sitting in a
warm IGM following beautifully the Hubble flow .
The idea is then to observe the Lyman forest of a number of QSOs, uniformly
distributed all over the sky, today and few years from now, and measure the shrink-
S. CRISTIANI ETC.
Fig. 2. – Upper panel: lhe transmission of the intergalactic medium at redshift z ∼ 4, observed
along the line of sight to the QSO Q0000-26. The average transmission is about 0.5 and hundreds
of features modulate the signal between 0 and 1. Lower panel: the expected shift in the spectral
features due to the cosmic deceleration in a time interval of 5 · 106yr.
ing/stretching pattern of the absorption features (see Fig. 2). The variation of the nor-
malized transmission is expected to be of the order of 10−6in ten years. Is it feasible?
Difficult as it may be, such an experiment is no more complex, nor more expensive,
nor of less fundamental importance than what our colleagues at CERN regularly do.
Accuracies not far from what we need for detecting the cosmic signal are presently
being reached in the observations of radial velocity perturbations induced by extra-solar
planets (e.g. HARPS ). We want to do the same but with objects that are hundred
thousand times fainter than the extra-solar planets targets, and on timescales of decades.
An extremely large light bucket is needed and in this respect the E-ELT is going to play
for European astronomers the role of LHC for particle physicists.
4. – The CODEX Team
On the basis of the conviction enunciated above, the CODEX (COsmic Dynamics
EXperiment) team has formed . From the work carried out by the team to explore
the feasibility of the experiment  we show here the results aimed at quantifying the
optimal redshift range and instrumental characteristics. They can be summarized by the
THE CODEX-ESPRESSO EXPERIMENT: COSMIC DYNAMICS, FUNDAMENTAL PHYSICS, PLANETS AND MUCH MORE...5
Fig. 3. – CODEX observing strategy assuming 2.2 nights/month of observation with a 42m ELT
over 15 years. The three different sets of data points represent different implementations of the
redshift drift experiment, each being optimal for a different goal: Blue points: 20 targets (in 4
bins), selected to give the highest overall radial velocity accuracy (2.13 cm/s). Yellow points: 10
targets selected to give the largest possible significance of a non-zero detection. Brown points:
2 targets, selected to give the best constraints on the acceleration and dark energy. The grey
shaded areas around the curves correspond to the present Ho uncertainty of +/−8 km/s/Mpc.
2?1 + zQSO
where the last exponent changes to −0.9 at zQSO> 4, due to the crowding of features
in the high-redshift Lyman forest. A resolution ℜ ≥ 30000 is sufficient to fully resolve
the Lyman features. A higher resolution ℜ ∼ 105would however be advisable in order
to better identify metal lines in the spectra. Different observing strategies are being
explored, according to different goals of the experiment (see Fig. 3) and it is reassuring
to know that already today there are at least 20 known QSOs with redshift between 2
and 5 bright enough to achieve a radial velocity accuracy of 3 cm/s in 3200 hours of
observation with a 42 m ELT.
Simulations have dictated the requirements for the spectrograph: a visible, high-
resolution spectrograph with an exquisite long-term stability. To achieve this goal with
an ELT is not trivial at all and requires an advanced design based on multiple modules
(presently five are envisaged) based on R4 echelle gratings.
6 Download full-text
S. CRISTIANI ETC.
A long list of potential problems is being investigated, including secular changes in the
structure of the IGM, the variability of the QSO continuum, weak lensing, the heliocentric
correction, instrumental issues such as the guiding accuracy, temperature control and the
scrambling of light in the fibers. None at the moment is recognized to be a show-stopper.
A special issue is represented by the wavelength calibration, because long-term sta-
bility is a must and serious doubts have been cast over the possibility to achieve the
required accuracy with conventional lamps. This is the reason why a new concept is
being developed: the laser frequency comb, an optical or near-IR laser generating a train
of femtosecond pulses with the pulse repetition controlled by an atomic clock, producing
a reference spectrum of evenly spaced δ functions 
Of course an instrument like CODEX would have many other applications, among
which it should be mentioned the variation of fundamental constants  to a precision
of one part over 108, or the detection of terrestrial extra-solar planets, in particular
following-up earth-mass candidates discovered through other techniques, and Big Bang
nucleosynthesis, with the determination of detailed primordial abundances in order to
clarify some of the present tensions existing for example about isotopic ratios such as
Is the leap from HARPS on the ESO 3.6m telescope in La Silla to the E-ELT too
daring?Maybe. This has prompted the CODEX Team to study the possibility of
building a precursor, dubbed ESPRESSO (Echelle Spectrograph for PREcision Super
Stable Observations) to be placed at the VLT, possibly at the incoherently combined
focus of the four UTs. The instrument would look like one CODEX module and would
allow the community to get a first glance of a significant part of the CODEX immediate
science, of course with the exception of the cosmic dynamics. ESPRESSO would make
it possible tests of the stability of the IGM that on the one hand are crucial for the
CODEX feasibility and on the other hand would provide fundamental information on
the formation of cosmic structures and feedback at high-redshift (see, for example ).
∗ ∗ ∗
A special thank to J.Liske, E.Vanzella and A.Grazian for their contribution to the
preparation of this article.
 Alpher R.A., Herman R.C., Reviews of Modern Physics, 22 (1950) 153.
 Cayrel R., Steffen M., Chand H. et al., A&A, 473 (2007) L37
 Levshakov S.A., Molaro P., Lopez S. et al., A&A, 466 (2007) 1077
 Liske J., Grazian A., Vanzella E. et al., MNRAS, submitted (2008) .
 Loeb A., ApJ, 499 (1998) L111.
 Lovis, C., Mayor, M., Pepe, F., et al. , astro-ph/0703024
 Miyoshi M., Moran J., Herrnstein J. et al., Nature, 391 (1995) 51.
 Murphy M.T., Udem Th., Holzwarth R. et al., MNRAS, 380 (2007) 839
 Pasquini L., Cristiani S., Dekker H., et al., Messenger, 122 (2005) 10.
 Penzias A.A., Wilson R.W., ApJ, 142 (1965) 419.
 Rauch M., ARA&A, 36 (1998) 267.
 Rauch M., Becker, G.D., Viel M., et al., ApJ, 632 (2005) 58.
 Smooth G., Bennet G.L., Kogut A., et al., ApJ, 396 (1992) L1.