Constraints on Omega_B, Omega_m and h from MAXIMA and BOOMERANG
Extracted source text ยท arXiv astro-ph/0010309
%\documentstyle[12pt]{article}
\documentclass[preprint,psfig]{aastex}
%\usepackage{aastexug}
\begin{document}
\title{Constraints on $\Omega_B$, $\Omega_m$ and $h$ from MAXIMA and BOOMERANG }
\author{T. Padmanabhan}
\affil{{Inter-University Centre for Astronomy \& Astrophysics, \\
Post Bag 4, Pune 411007, India} }
\email{paddy@iucaa.ernet.in}
\and
\author{Shiv K. Sethi}
\affil{Mehta Research Institute,\\ Chhatnag Road, Jhusi, Allahabad
211 019, India}
\email{sethi@mri.ernet.in}
\begin{abstract}
We analyse the BOOMERANG and MAXIMA results in the context of
models with $\Omega_{\rm Total} = 1$ and $n_s =
1$. We attempt to constrain three other parameters---$h$, $\Omega_B$, and
$\Omega_m$---from these observations. We show that: (a)
the value of $\Omega_B h^2$ is too high to be compatible with
primordial nucleosynthesis observations at 95\% confidence level (b)
universe with age greater than
$ 12 \, \rm Gyr$ is ruled out at 95\% confidence level and
(c) the value of $\Omega_m h$ is too high
to be compatible with the shape of the power spectrum of
gravitational clustering. In effect, our analysis shows that
models with $\Omega_{\rm Total} = 1$ and $n_s = 1$ are
ruled out by BOOMERANG and MAXIMA observations.
\end{abstract}
Precise determination of CMBR anisotropies has long been expected to
give accurate values of cosmological parameters (see e.g. Bond,
Efstathiou \& Tegmark 1997 and references therein). These
cosmological parameters include parameters of background FRW model
($\Omega_{\rm Total}$, $\Omega_\Lambda $, $h$, $\Omega_B$, etc.),
parameters that determine the formation of structure in the universe
( $\sigma_8$, scalar spectral index $n_s$, etc.), and
the parameters related to the re-ionization of the universe (the optical depth to
the last scattering surface, $\tau$ etc.).
While the future
experiments MAP and Planck \footnote{For details see
{\tt http://map.gsfc.nasa.gov} and
{\tt http://astro.estec.esa.nl/SA-general/Projects/Planck}},
largely owing to their all-sky coverage,
are expected to determine most of these parameters with a few percent
accuracy (Jungeman {\it et al. \/} 1996, Zaldarriaga {\it et al.}
1997, Prunet {\it et al.} 1999), recent observations have already begun to give important
clues about some of these parameters (Miller {\it et al. \/} 1999,
Mauskopf {\it et al.} 2000, Netterfiled {\it et al.} 1997, for a
summary of observation up to 1998 and parameter estimation from these
observations, see Lineweaver \& Barbosa 1998). Recent balloon experiments
BOOMERANG and MAXIMA reported CMBR anisotropy measurements at
angular scales between $\simeq 10^\circ$ and $ \simeq 10'$ (de Bernardis {\it
et al. \/} 2000, Hanany {\it et al. \/} 2000). These
experiments observed nearly $1 \%$ of the sky with angular
resolution $\simeq 10'$. For both these experiments the cosmic
variance was small enough (owing to the sky coverage)
to determine precisely the position of the first
Doppler peak ($\ell \simeq 200$) of the CMBR anisotropies. Both
BOOMERANG and MAXIMA results gave strong evidence that
$\Omega_{\rm Total} = 1$ (de Bernardis {\it
et al. \/} 2000, Hanany {\it et al. \/} 2000), which was already
indicated by other observations (Netterfield {\it et al.} 1997).
While the position of the first Doppler peak give unambiguous evidence about
the geometry of the universe, determination of other cosmological
parameters is more difficult. This is because variation in several
different parameters give the same change in measured anisotropies,
e.g. the height of first Doppler peak is nearly degenerate in $\Omega_B$,
$h$, $\Omega_\Lambda$ and $n_s$. Some of this degeneracy can be
lifted with the measurement of anisotropies at even smaller angular
scales. BOOMERANG and MAXIMA probe with angular scales
corresponding to multipoles $\ell_{\rm max} \simeq
\{600,700\}$, respectively, which is up to or beyond the
expected position of the second Doppler peak. Though the results of
these experiments have not been able to find the position of the
second Doppler peak, accurate measurement of anisotropies at such angular
scales is expected to break some of the degeneracy which measurements
near the first Doppler peak alone cannot.
The BOOMERANG and MAXIMA data have been used to determine various
cosmological parameters (Balbi {\it et al. \/} 2000, Jaffe {\it et
al. \/} 2000, Lange {\it et al.} 2000, Tegmark \& Zaldarriaga
2000). Combined with other independent measurements
of cosmological parameters
(e.g. measurement of $\Omega_B h^2$ from element abundance,
measurement of $h$ from nearby observations or inference about the
values of $\Omega_\Lambda$ and $\Omega_m$ from the SN1a data, etc.)
these data are expected to lead to a unique picture. However, owing
to degeneracies in parameter estimation, the value of estimated
parameters and their errors depend sensitively on various assumption
related to the assumed allowed range of parameters, i.e. on the
priors on the parameters.
In this letter, we perform a likelihood analysis on the band-powers
reported by the MAXIMA and BOOMERANG experiments. However, instead of
of doing a multiple parameter analysis, we fix the values of most
parameters from other considerations (and prejudices!) and then attempt to
estimate just three parameters---$\Omega_B$, $h$, and
$\Omega_m$ assuming weak
priors on their allowed values. In the next section, we explain our
choice of parameters and the method we use in brief. In the third
section, we present and summarise our results.
\section{Cosmological parameters and CMBR data}
Most generic models of inflationary scenario give two unique
predictions: $\Omega_{\rm Total} = 1$ and $n_s \simeq 1$
(see e.g. Steinhardt 1995). The first
of these predications is confirmed by BOOMERANG and MAXIMA. The
analysis of COBE-DMR data is consistent with $n_s = 1$ (Bennett {\it
et al \/} 1996, Bunn \& White 1996). Therefore it
is reasonable to believe that the current data is in good agreement
with these predictions.
We fix the value of these two parameters based on these
considerations and use $\Omega_{\rm Total} = 1$ and $n_s \simeq 1$
throughout. (It should be pointed out that
$\Omega_{\rm Total} = 1$ is a stricter requirement of inflation than
$n_s = 1$; one gets $n_s = 1$ only for exponential inflation; see e.g. Steinhardt
1995.) Also note that we are not concerned with the origin of the values
$\Omega_{\rm Total} = 1$ and $n_s =
1$. For example, these could arise merely from the requirement of
scale invariance
for the background universe (giving $\Omega_{\rm Total}=1$) and the
perturbations (giving $n_s=1$) without invoking inflation --- as was
originally done by Harrison and Zeldovich, years before inflation was
invented (Harrison 1970, Zeldovich 1972).
But, of course, inflationary models made these parameter values fairly
well accepted.
Other parameters like $\Omega_\Lambda$, $h$ and
$\Omega_B$ cannot be fixed by theoretical considerations alone.
In our analysis we assume a non-zero $\Omega_\Lambda$ because
recent high-z SN1a observations suggest a non-zero cosmological constant
(Perlmutter {\it et al.} 1999, Riess {\it et al.} 1998).
We do not consider CMBR anisotropies from tensor perturbations as
most models of inflation give negligible contribution from tensor
perturbations for $n_s =1$ (Steinhardt 1995).
Re-ionization of universe can also alter
primary CMBR anisotropies significantly. Present observations suggest
that the universe is ionized up to $z \simeq 5$. For the CMBR
anisotropies to be significantly altered by re-ionization, the
minimum re-ionization redshift should lie between $10$ and $20$ (see
e.g. Bond
1996). Therefore, keeping the constraints from present observations in
mind, we neglect the effect of re-ionization on the measured CMBR
anisotropies.
We fix the value of $\Omega_m + \Omega_\Lambda = \Omega_{\rm
Total} = 1$, and compute the confidence levels on the best-fit
values of $\Omega_B$ and $h$, and $\Omega_m$
from MAXIMA and BOOMERANG observations.
The range of parameters in which the minimum of $\chi^2$ is
searched is: $0.01 \le \Omega_B \le 0.15$, $0.4 \le h \le 1.1$ and
$0.1 \le \Omega_m \le 0.95$.
The $\chi^2$ for the model comparison with observations is given by:
\begin{equation}
\chi^2 = \sum_{i = 1}^N \left ({{\cal C}_\ell^{\rm obs} - {\cal
C}_\ell^{\rm th} \over \Delta {\cal C}_\ell^{\rm obs}}
\right)
\label{eq:a1}
\end{equation}
Here $N = 22$ (10 points from MAXIMA and 12 points from BOOMERANG),
${\cal C}_\ell^{\rm obs}$ are the measured band-powers and
${\cal C}_\ell^{\rm th}$ are the theoretical band-powers, $\Delta
{\cal C}_\ell^{\rm obs}$ are the errors on measured band-powers.
We do not take into account the calibration uncertainties in our
analysis. The
theoretical band-powers are calculated using the CMBR Boltzmann code
CMBFAST (Seljak \& Zaldarriaga 1996), which gives COBE-normalized
(normalized according to the fitting formula of Bunn \& White
(1996)) angular power spectra.
Eq.~(\ref{eq:a1}) assumes that different band-powers are
uncorrelated. This assumption is valid only if the anisotropies are
measured over the entire sky (see e.g. Peebles 1993, Padmanabhan 1993).
We assume lack of correlation in this work also
because the covariance matrix of band-powers has not been made
public yet.
Eq.~(\ref{eq:a1}) implicitly assumes that the
likelihood function is Gaussian in band-powers near its
maximum. While this assumption is true in principle, in practice
there can be significant deviation from Gaussianity near the
maximum. Bond, Jaffe, and Knox (2000) advocate using another variable
instead of band-powers for doing the maximum likelihood
analysis. We do not use it here.
However, while quoting errors we do not use the Fisher matrix approach
which can give meaningful results only for the Gaussian case.
Instead we directly give
the confidence levels on $\Delta \chi^2$.
\section{Results}
Our results are shown in Figures~1 and~2. The $\chi^2$ for 22 data
points from BOOMERANG and MAXIMA with three fitted parameters
(i.e. 19 degrees of freedom) is 22.4, which is an
excellent fit (Goodness-of-fit probability, $Q = 0.26$). The best fit values
and $1\sigma$ errors are:
$\Omega_B = 0.075^{{+}0.003}_{{-} 0.015}$, $h =0.62^{{+}0.08}_{{-}
0.02}$ and $\Omega_m= 0.86^{{+}0.04}_{{-} 0.21}$. The best fit
model along with BOOMERANG and MAXIMA data points is shown in
Figure~3.
The range of allowed $h$ is in fair agreement with the
measurement of $h$ from local observations (for a recent summary of
these results, see Primack 2000). Recent SN1a suggest that that the
universe is flat with $\Omega_\Lambda + \Omega_m = 1$ with
$\Omega_m = 0.28^{{+}0.09{+}0.05}_{{-}0.08{-}0.04}$ (1$\sigma$) .
This is within $\Delta \chi^2 = 2$ of the value inferred by our
analysis.
In Figure~1 we show the confidence levels in the
$\Omega_B\hbox{--}h$ plane. The region bounded by the contours
correspond to $\Delta \chi^2 \le 2.3$ and $\Delta \chi^2 \le 6.17$, which, for Gaussian errors,
corresponds to 68~\% (1$\sigma$) and 95.4~\% (2$\sigma$) for
two-parameter fits. We also show the 95\% region allowed by
primordial nucleosynthesis observations (for a recent review see
Tytler {\it et al} 2000). As seen in
the figure, the region allowed by CMBR observations is at variance
with the predictions of primordial nucleosynthesis at at least 95\% level. This
result is in agreement with other analyses on the same data set
(Tegmark \& Zaldarriaga 2000, Jaffe {\it et al} 2000).
The region corresponding to one and two $\sigma$ in the
$\Omega_m\hbox{--}h$ plane is shown in Figure~2. Our results
show that the current CMBR
observations favour a universe with age $\le 11 \rm Gyr$ and
are incompatible with a universe of age $\ge 12 \, \rm Gyr$ at
$\simeq 95 \%$ level. Though this is on the lower side of the
expected age of the universe from estimated ages of globular clusters,
etc., this is not in disagreement with those observations (for a
recent status report see Primack 2000).
Another important constraint on the values of $\Omega_m$ and $h$ comes
from the shape of the power spectrum of galaxy clustering (see
e.g. Bond 1996). These observations, within the context of a flat
model with cosmological constant, give $\Omega_m h \simeq 0.25 \pm
0.05$. We plot this region in the $\Omega_m\hbox{--}h$ plane in
Figure~2. If we require that $h \le 0.75$, as most recent
measurements of $h$ suggest (Primack 2000), then CMBR observations
exclude the region required to satisfy the galaxy clustering
observations by more than $95\%$ confidence level.
In conclusion, recent BOOMERANG and MAXIMA observations, within the
context of simplest inflationary models ($\Omega_{\rm Total} = 1$ and
$n_s \simeq 1$),
imply: (1) too large a of value
of $\Omega_Bh^2$ to be compatible with primordial nucleosynthesis
observations, (2) an age of universe $t_0 \le 12 \, \rm Gyr$, and (3)
too large a value of $\Omega_m h$ to be in agreement with the shape of
the power spectrum of galaxy clustering. In view of this, it is
safe to conclude that
the models with $\Omega_{\rm Total} = 1$ and $n_s = 1$ are
ruled out by these observations.
\begin{thebibliography}{}
\bibitem{} Balbi, A. {\it et al. \/} 2000, astro-ph/0005124
\bibitem{} Bennett, C. L. {\it et al. \/} 1996, ApJ, 464, L1
\bibitem{6} Bunn E. F. \& White M. 1996, astro-ph/9607060
\bibitem{} Bond, J. R., Efstathiou, G., \& Tegmark, M. 1997, MNRAS,
291, L33
\bibitem{} Bond, J. R., Jaffe, A. H., \& Knox, L. 2000, ApJ, 533,
19
\bibitem{} Bond, J. R. 1996, Observations of Large-Scale Structure
in the universe, Ed. Schaeffer, R., Les Houches Lectures,
Elsevier Science Publishers
\bibitem{} de Bernardis, P. {\it et al. \/} 2000, Nature, 404, 995
\bibitem{} Hanany, S. {\it et al. \/ } 2000, astro-ph/0005123
\bibitem{} Harrison, E. R. 1970, Phys. Rev. D, 1, 2726
\bibitem{} Jungman G., Kamionkowski M., Kosowsky A., Spergel D. N. 1996,
Phys. Rev. D, 54, 1332
\bibitem{} Lange, A. E. {\it et al. \/ } 2000, astro-ph/0005004
\bibitem{} Lineweaver, C. H. \& Barbosa, D. 1998, ApJ, 496, 624
\bibitem{} Mauskopf, P. {\it et al. \/ } 2000, ApJ, 536, L63
\bibitem{} Miller, A. D. {\it et al. \/} 1999, ApJ, 524, L1
\bibitem{} Netterfield, C. B., Devlin, M. J., Jarosik, N., Page,
L. \& Wollack, E. J. 1997, ApJ, 474, 47
\bibitem{} Padmanabhan, T. 1993, {\it Structure Formation in
the Universe}, Cambridge University Press
\bibitem{} Peebles, P. J. E. 1993, {\it Principles of Physical
Cosmology}, Princeton university Press
\bibitem{} Perlmutter, S. {\it et al. \/} 1999, ApJ, 517, 565
\bibitem{} Primack, J. 2000, astro-ph/0007187
\bibitem{} Prunet, S., Sethi, S. \& Bouchet, F. 1999, astro-ph/9911243
\bibitem{} Riess, A. {\it et al.} 1998, AJ, 116, 1009
\bibitem{} Seljak, U. \& Zaldarriaga, M. 1996, ApJ, 469, 437
\bibitem{} Steinhardt, P. J. 1995, IJMP A, 10, 1091
\bibitem{} Tegmark, M. \& Zaldarriaga, M. 2000, astro-ph/0004393, to appear
in PRL
\bibitem{} Tytler, D., O'Meara, J. M., Suzuki, N, \& Lubin, D. 2000,
Physica Scripta, T 85, 12
\bibitem{} Zeldovich, Ya. B. 1972, MNRAS, 160, 1p
\bibitem[] Zaldarriaga M., Spergel D. N., Seljak U. 1997, ApJ, 488, 1
\end{thebibliography}
\newpage
\begin{figure}
\figurenum{1}
%\plotone{fig1.ps}
\includegraphics[angle=270, width=\textwidth]{fig_homegab.ps}
\caption{The contours correspond to allowed 1 and 2 $\sigma$
regions by CMBR observations (see text for detail). The cross-hatched region
correspond to the 95\% region ($\simeq 2\sigma$)
from primordial nucleosynthesis (Tytler {\it et al.} 2000). \label{fig:1}}
\end{figure}
\begin{figure}
\figurenum{2}
%\plotone{fig1.ps}
\includegraphics[angle=270, width=\textwidth]{fig_homegam.ps}
\caption{The contours correspond to allowed 1 and 2 $\sigma$
regions by CMBR observations (see text for detail). The hatched
region corresponds to $\Omega_m h = 0.25 \pm 0.05$. Other curves show the
age of the universe $t_0$ (value indicated in the figure legend)\label{fig:2}}
\end{figure}
\begin{figure}
\figurenum{3}
%\plotone{fig1.ps}
\includegraphics[angle=270, width=\textwidth]{fig_bestfit.ps}
\caption{The best fit model is plotted along with BOOMERANG and MAXIMA data
points, which correspond to empty and filled polygons, respectively.}
\end{figure}
\end{document}