< home

cjo25

CMB

Summary and figures of the currently available experimental data on CMB temperature anisotropies. A couple of packages for download, including a C version of the WMAP likelihood code.

NEWS

previous news
  • 10.07.03: I am now in Switzerland. News and features on this page.
  • 01.07.03: New 'Talks and posters' page here with abstracts and downloads of slides and posters in PDF format.
  • 01.07.03: I have come back from Wales and I have put the slides of my seminar here. I am on my way to Switzeland to write up and any other plans can be found here.
  • 28.05.03: Less cluttered navigation. Publications now available here. Also, I have added an online research proposal here.
  • 16.05.03: Another update, another design :). I have now written my WMAP page with a C version of the likelihood code for download and some recent results.

Publications

List of publications with abstracts and links.

Talks and posters

List of talks an posters with abstracts and downloads.

Contact: Carolina Ödman


Research Proposal: Quantitative Cosmology

download: PS PDF

1 - Introduction

In recent years various cosmological experiments have reached unprecedented accuracy. Cosmic Microwave Background (CMB) data provide unique observations of the early universe with the recent WMAP result [1] and other high resolution measurements [2]. Large redshift surveys such as the 2dFGRS [3] and the Sloan digital sky survey [4] have already made available measurements of about 150,000 galaxies each at redshifts z <= 0.45. Distant supernove (SNe) data [5], [6] and other cosmological observations such as Lyman-α forests [7] and quasar surveys [8] complete this history of the universe at low and intermediate redshifts.

Such observations are powerful probes of cosmological theories from the very early universe through structure formation until the present state of the universe. Subsequently, cosmological parameter estimation has been a growing field and combinations of observations have enabled us to break a number of degeneracies inherent to each probe. As a result, an increasingly consistent image of a rather simple universe arises. The universe is believed to be flat, dominated by dark energy and dark matter and observations suggest primordial fluctuations have properties naturally occurring in inflationary scenarios.

These inferences rely on prior assumptions about cosmological models, certain theoretical approximations and the understanding of experimental uncertainties. Cosmological parameters are now estimated to such precision that higher order effects should be accounted for in a complete analysis.

I propose to investigate this aspect of cosmological parameter estimation in three ways. In section 2, I propose to improve and complete present methods of parameter estimations. In section 3 I propose to test data sets against each other using phenomenological fits and by other means for a better understanding of the experimental data. Finally in section 4 I propose to use this general method to find areas of theoretical uncertainty in order to focus on particular aspects of cosmological theories which are not yet fully understood or approximated in the present approach. I conclude by summarising.

2 - Method for exploring large parameter spaces

Cosmological parameter estimation is traditionally approached from a Bayesian point of view and several methods have been developed including grids of models and iterative processes such as Markov Chain Monte Carlo codes [9]. With the data becoming increasingly accurate, the usual approach can lead to prohibitive computational cost. It is therefore timely to make use of complementary methods for a first estimate, followed by a more detailed analysis of a smaller regions of the cosmological parameter space.

I propose to perform a general Fisher information analysis on very large parameter spaces. I then propose to carry out a principal component analysis on this result. This will enable me to explore the full cosmological parameter space effectively with sampling methods making use of the principal components. As a by-product of this, I will be able to extract degeneracies between parameters which I will then investigate by analysing higher moments. This will be useful for refining theoretical models as mentioned in section 3 and provide a basis for semi-analytic approaches for testing data sets, as I describe in section 4.

I am presently following the method outlined above to explore the effect of the inclusion of correlated isocurvature initial conditions in a CMB-based parameter estimation. This adds about ten dimensions to the cosmological parameter space and requires such an approach.

3 - Experimental data, uncertainties and systematics

Combining data sets is the key to breaking degeneracies among parameters inherent to each type of observation. A caveat of this approach is that experimental errors add up, as in any manipulation of experimental data. Therefore, I propose to investigate experimental errors and their effect when possible.

I have already carried out such an analysis using a phenomenological approach to the CMB anisotropy power spectrum and have found that CMB experiments using different instruments and in different frequency ranges give consistently different results [10]. Bolometer experiments at high frequency suffer from uncertainties that tend to smear the CMB power spectrum and predict more baryonic matter than interferometer measurement at lower frequencies. This result has convinced me of the importance of systematics and has led me to believe that this approach should be pursued.

I propose to build phenomenological models to fit CMB observations and constrain them with subsets of data in order to examine and evaluate the signature of experimental systematics. I shall then compare those results with the discrepancies in cosmological parameter estimations from the same subsets of data in order to evaluate the effect of experimental uncertainties.

4 - Refining the theoretical models

Although the general picture is consistent, the details of cosmic evolution are still to be understood. Reionization of the universe, dark energy, inflation and the end of inflation (reheating) are major gaps in our cosmological theory. Although future experimental results will allow us to start constraining such models, they will also require us to improve our theoretical framework.

I propose to make use of results from investigation of the present cosmological parameter space to work out critical directions for more focused research. Finding degeneracies among cosmological parameters and where the models fail to fit the data, such as the WMAP power at low multipoles of the TT and TE power spectra are useful starting points. I therefore propose as a longer term project to focus on those grey regions of our models in order to characterize and refine them.

In particular, I propose to test theories of dark energy and to devise a set of dark energy parameters which describe it without being confined to a class of models. This would be a follow-up of a project I have been working on with A. Melchiorri and others [11]. Similarly, I would like to use new data to test whether the present parameters assumed for describing inflation are appropriate (see [12] and references therein). Upcoming results from ongoing and future experiments will provide new means of constraining properties of both inflation and dark energy.

Summary

By developing a powerful cosmological parameter estimation pipeline making use of several complementary methods, I propose to investigate both experimental data for uncertainties and cosmological theories. This approach will allow me to extract certain regions of the current cosmological parameter space for more detailed investigation. Future cosmological data such as CMB polarization and Sunyaev-Zel'dovitch measurements and Lyman-α observations will provide new ways of testing increasingly detailed cosmological models.

Bibliography

  1. D. N. Spergel & al., astro-ph/0302209
    L. Verde & al., astro-ph/0302218
  2. Kuo & al., astro-ph/0212289
    Pearson & al., astro-ph/0205388
    Grainge & al., MNRAS, 341, 4, L23
  3. http://msowww.anu.edu.au/2dFGRS/
  4. http://www.sdss.org/
  5. S. Perlmutter & al., Astrophys. J, 517 (1999) 565-586
  6. B. Schmidt & al., Astrophys. J, 507 (1998) 46-63
  7. R. Croft & al., Astrophys. J, 520, 1 (1999)
  8. http://www.2dfquasar.org/
    http://archive.stsci.edu/sdss/quasars/
  9. A. Lewis, S. Bridle, Phys.Rev. D66 (2002) 103511
  10. C. Ödman & al., Phys. Rev. D.67 (2003) 083511
  11. A. Melchiorri & al., astro-ph/0211522
  12. A. Liddle, S. Leach, astro-ph/0305263

Previous projects:

The shape of the CMB:
first results : Phys. Rev. D67 083511 - Collaboration with Alessandro Melchiorri Anthony Lasenby andn Mike Hobson
VSA analysis : MNRAS, 341, 4, 1084 - MNRAS, 341, 4, L23 - for the VSA team at MRAO
Poster presented at Joe Silk's birthday conference, Oxford, December 2002
post-WMAP results : Proceedings of the 2nd CMBNet Meeting, Oxford, Feb 2003

Gravitational Waves:
Phys. Rev. D 67, 021501(R) - Collaboration with Alessandro Melchiorri when the Archeops CMB observations became available.

Dark Energy:
Seminar given at the University of Wales, Swansea, June 2003
Talk given at the Royal Astronomical Society, London, March 2003
astro-ph/0211522 (PRD, submitted ) - collaboration with Alessandro Melchiorri, Laura Mersini and Mark Trodden
Proceedings of the XVIII IAP Colloquium "On the nature of dark energy" Paris, July 2002 - collaboration with Alessandro Melchiorri