Research Interests

My research has been focused on aspects of weak gravitational lensing as a probe of cosmology. If you are unfamiliar with the concept of weak lensing you can read my brief introduction to weak lensing.

Constraints on Cosmology

The basic premise is that large scale structure in the Universe (i.e. structure in the dark matter distribution) will induce small changes in the observed shapes of galaxies. A statistical measure of this distortion can probe the matter distribution on large scales. Specifically it can be related to the projected matter power spectrum, which can be modeled based on theory, hence we can assess which cosmological model best fits the observed matter power spectrum. In doing this we provide constraints on cosmological parameters, most notably the density of matter in the Universe (Ωm) and the variance of the mass density on large scales (σ8).

My master's work brought together data from 4 surveys to constrain cosmological parameters, the result is the largest such study ever performed, consisting of 100 square degrees. This unprecedented data set together with some important amendments to the analysis yield constraints on Ωm and σ8 that greatly reduce previous tension with the results from WMAP.

Tomography and Photometric Redshifts

The weak lensing I have discussed thus far does not make use of each galaxies redshift, though it is crucial that the distribution of galaxies as a function of redshift is known it is not important which galaxy has what redshift. In essence we integrate the measured signal over the redshift distribution so as long as the global properties of the distribution are known we need not concern ourselves with the details.

Tomography attempts to gain information by measuring the weak lensing signal at different redshifts. This provides information on how the cosmological parameters are evolving with time in the Universe. This is critical if weak lensing is going to provide constraints on dark energy.

The difficulty is that one needs to have a measured redshift for every galaxy in order to split the sample into redshift slices. Spectroscopic redshifts are very hard to obtain for faint galaxies, and weak lensing studies require many millions of galaxies to accurately measure the small distortions in shape. As a practical matter it is impossible to use spectroscopy to obtain redshifts for millions of galaxies, thus we must use the less accurate method of photometric redshift determination.

The field of photometric redshifts is still relatively young and there are many different techniques available. Unfortunately it has been difficult to assess the relative accuracy of the many algorithms and nuanced analyses that have been developed. Hopefully this will improve through a community wide initiative known as PHAT (Photometric redshift Accuracy Testing). This program aims to test the various methods on sets of simulated and real data in order to improve these methods and to understand their accuracy.

My current work is geared in this direction. To help understand and quantify the level of contamination between photometric redshift bins I have been developing an analysis technique to measure the contamination via the angular correlation function. The goal for weak lensing is to identify which photometric redshift rage has the least contamination, and thus is most suitable to use for the purposes of tomography. Additionally the inevitable small level of contamination can be assessed and this error can be taken into account in a weak lensing analysis.

 
 

Research: