I work on a wide variety of topics in cosmology and extragalactic astronomy.  Much of my work concerns precision cosmological constraints from the measurement of the large-scale structure of the Universe.  Another major theme is the evolution of galaxies and their relation to their environment.  My work spans observations, theoretical astrophysics, and numerical simulations.

Please see my bio for recent focus areas.  Key collaborations presently include the Dark Energy Spectroscopic Instrument collaboration, the James Webb Space Telescope Near-Infrared Camera team, the Euclid Consortium, and our Abacus N-body code.

I previously served as the Director of the Sloan Digital Sky Survey III, which includes a major large-scale structure known as the Baryon Oscillation Spectroscopic Survey, as well as surveys of the stellar population of the Milky Way galaxy and a search for extra-solar planets.  


The Sloan 2.5-m Telescope at Apache Point Observatory

Cosmology Research

A central focus of my work has been the development of the baryon acoustic oscillation method for the study of the cosmological distance scale and the evolution of dark energy.  My collaborators and I have studied the behavior of the acoustic peak in both linear perturbation theory and numerical simulations and the practical implementation of the method on real data.  I have been involved in many of the largest survey projects in astronomy, most notably the Sloan Digital Sky Survey.

Currently I am the Director of the Sloan Digital Sky Survey III, which includes a major large-scale structure known as the Baryon Oscillation Spectroscopic Survey.  SDSS-III/BOSS will soon produce the most stringent cosmological tests from any redshift survey.  Check this space!

I continue to work on SDSS-II data as well.  Our 2005 paper announcing the discovery of the acoustic peak in the correlation function of SDSS data led to the development of a web page with various explanations and illustrations of the baryon acoustic oscillation phenomenon.  Our most recent papers in 2012 achieve a 2% measurement of the distance to redshift 0.35; we infer from this a Hubble constant of 70 +- 2 km/s/Mpc and Omega_m of 0.27 +- 0.015!

In recent years, the acoustic peak method has become one of the mainstays of the study of dark energy.  I have been involved with several different survey concepts, including ADEPT/JDEM/WFIRST and the Gemini/Subaru Wide-Field Multi-Object Spectrograph (which has now evolved into the Prime Focus Spectrograph concept).  

My group has been pushing to demonstrate the reliability of the acoustic peak method down to 0.1% in distance, the statistical limit of the most aggressive of these surveys.  We have developed large suites of cosmological simulations to achieve this.  We are now working on the development of a new generation of cosmological N-body simulation code.

The evolution of a point-like initial density perturbation from Eisenstein, Seo, and White (2007).  The graph shows the motion of the four cosmic components in spherical comoving coordinates.  The baryon-photon acoustic wave travels outward to reach a radius of 150 Mpc (500 million light-years) and then stalls, creating an excess in the clustering of galaxies that we can still detect today.  More information about this animation.

Baryon Acoustic Oscillations

In the first million years after the Big Bang, the cosmic plasma rings with sound waves excited by the initial inflationary(?) perturbations.  The radiation pressure from the cosmic microwave background keeps the ionized gas from clustering; this pressure leads to relativistic sound waves that propagate until the Universe becomes neutral at redshift 1000.  These sound waves lead to the dramatic acoustic oscillations seen in cosmic microwave background anisotropy data, but they also leave a faint imprint in the clustering of galaxies and matter today.  The signature is a small excess in number of pairs of galaxies separated by 150 Mpc (500 million light-years).  These features are often referred to as the baryon acoustic oscillations (BAO).

My group uses this acoustic peak to measure the cosmic distance scale and thereby probe the acceleration of the expansion history of the Universe and the properties of dark energy.  The acceleration causes higher redshift objects to appear more distant.  We can measure distance by determining the angle subtended by something of known size; in this case, we are using the peak in galaxy correlations at 150 Mpc as the feature of known size.

In 2005, I led a group of researchers from the Sloan Digital Sky Survey (SDSS) in the detection of the acoustic peak in the luminous red galaxy (LRG) sample.  This produced a 4% measurement of the distance to redshift 0.35.  Since then, the BAO method has been accepted as one of the major pieces of our study of dark energy.  My collaborators and I have been pursuing the acoustic peak with larger samples in SDSS and SDSS-III, pushing the state of the art in the theory of the BAO and its extraction from large data sets, and designing the next generations of large redshift surveys.

Further pedagogical material about the BAO method

We prepared a web site about the 2005 detection paper.  In particular, this site includes an illustrated description of the acoustic peak and a non-technical description.  More technical descriptions can be found in this version and in Eisenstein, Seo, & White (2007).

Chuck Bennett and I wrote a Physics Today article about the BAO in 2008.

Weinberg et al. (2012) presents a major review of the observational probes of dark energy, including a chapter on the BAO method.  This is for those looking for a presentation at the introductory graduate level.  Eisenstein & Hu (1998) also contains helpful orientation about the BAO.

I have given many lectures on the BAO, some of which are on-line.  For example, here are my Loeb Lectures from fall 2009.

Observations of the Baryon Acoustic Oscillations

My group has been involved in a series of papers using SDSS I and II data to study the BAO.

Our first detection was in Eisenstein et al. (2005) using SDSS DR3 and the Luminous Red Galaxy sample.  We used the galaxy correlation function to detect the acoustic peak and measure the distance to z=0.35 to 4%.

We have continued this work into later data releases (DR4, DR5, and DR7) and using both power spectrum and correlation function analysis.  These were published as Tegmark et al. (2006), Percival et al. (2007a), Percival et al. (2007b), Okumura et al. (2008), Percival et al. (2010), and Reid et al. (2010).  This culminated in a 2.7% measurement of the distance to z=0.275.

Our most recent work adds the methods of density-field reconstruction.  With this, we achieve a 1.9% measurement of the distance to z=0.35.  Reconstruction is a method for combining the observed galaxy distribution with the theory of gravitational structure formation so as to better estimate the initial large-scale density perturbations.  Our work is now submitted as Padmanabhan et al. (2012), Xu et al. (2012), and Mehta et al. (2012).  The data files and figures from this work are available.

I am currently Director of SDSS-III; here is a summary of the project.  The SDSS-III Baryon Oscillation Spectroscopic Survey will soon produce the best BAO results to date, with the goal of reaching a 1% measurement of the cosmic distance scale from BAO.

Theory of the Baryon Acoustic Oscillations

The BAO phenomena rests on simple physical ideas from the early Universe.  However, the manifestation in low-redshift data is complicated by non-linear structure formation, including redshift distortions and galaxy clustering bias.  My collaborators and I have sought to validate the BAO method at the 0.1% level, approximately the statistical errors available to a cosmic variance limited survey.  Much of this theory is summarized in Weinberg et al. (2012).

My early work, such as Eisenstein & Hu (1998) and Eisenstein, Hu, & Tegmark (1998), explored the BAO as a standard ruler and predicted that it would be usefully detected in data sets the size of SDSS.

In Seo & Eisenstein (2003) and Seo & Eisenstein (2007), we construct Fisher matrix predictions for the distance scale and dark energy measurements that could result from large redshift surveys.

In Seo & Eisenstein (2005), we conducted our first investigation of the non-linear effects on the BAO, using a suite of N-body simulations.  We continued this with better simulations and more detailed analysis in Seo et al. (2008), Seo et al. (2010), and Mehta et al. (2011).  Our latest results argue that the acoustic scale in the matter field is stable to 0.02%, while the accuracy in the galaxy field can reach 0.1% (the limit of our current simulations).  Xu et al. (2012) contains further demonstrations of this in even more realistic mock catalogs from the LasDamas collaboration.

We give physical arguments for the robustness of the acoustic peak in Eisenstein, Seo, & White (2007).  This picture then led to the suggestion of density-field reconstruction in Eisenstein, Seo, Sirko, & Spergel (2007).  Reconstruction has been tested in several papers, including the N-body papers listed above, and now performed on data in Padmanabhan et al. (2012).

Pat McDonald and I investigated the detectability of the BAO in the Lyman alpha forest in our 2007 paper.  This idea is now being pursued by the SDSS-III BOSS survey, using a grid of over 100,000 high-redshift quasars.

Galaxy Evolution Research

I have been active in the Sloan Digital Sky Survey since 1998.  I was the lead developer of the Luminous Red Galaxy sample, which led to my interests in the evolution of red galaxies.  I have been involved with studies of the galaxy luminosity function, the color and detailed spectral properties of red galaxies, and the correlations of galaxy properties with environment.

The SDSS Luminous Red Galaxy sample has been extended by the 2dF-SDSS LRG and Quasar (2SLAQ) survey and now by the SDSS-III Baryon Oscillation Spectroscopic Survey.  SDSS-III BOSS has already measured the redshifts of over 500,000 luminous red galaxies!

I have been involved with two field galaxy redshift surveys that pushed substantially fainter than SDSS.  In the AGN and Galaxy Evolution Survey (AGES), we used the MMT to survey 8 square degrees to a depth of I=20.  Then in the PRIsm MUltiobject Survey (PRIMUS), we used the Magellan Telescope to survey 9 square degrees as deep as i=23.  Both surveys were conducted in regions with excellent multi-wavelength imaging data.  We have used the data to investigate the evolution of galaxies from z=1 as well as the properties and evolution of active galaxies. 

At the highest redshifts, I was a member of the NICMOS Ultra-Deep Field team.  I am the extragalactic theme lead for the James Webb Space Telescope Near-Infrared Camera instrument team.

An image of the raw data from the PRIMUS survey.  Each tiny streak is the spectrum of a faint galaxy; this single exposure contains over 2000.  The large dark circles are the result of a lack of targets near bright stars.