Research

Overview

My scientific work broadly revolves around contemporary questions in cosmology as well as the physics of galaxies. Within this context I am keen on finding out more about

cosmological parameters that quantify the average mix of various forms of matter and energy throughout the Universe;
the rate of space expansion between galaxies in the past;
the large-scale distribution of (dark) matter, in particular about the distribution between and around galaxies;
the statistical difference between the spatial matter and galaxy density, the so-called galaxy bias;

My main observational tool for addressing these topics is the (weak) gravitational lensing effect -- a relatively new exciting method that has become viable only about a decade ago.

The general picture

The huge success of the standard cosmological model in explaining a wide range of cosmological observations makes it the currently favoured model for cosmologists (LCDM; a cold dark matter model with cosmological constant Lambda). This model explains the global expansion of the cosmos, which is now known to be accelerating over the recent several billion years, and the formation of structure as natural consequence of gravitational attraction. Basically, we picture the history of the cosmos as an evolution process over about 13.8 billion years from a simple (homogeneous), very hot, featureless state into the structured and cold Universe that we can see today. Objects, such as stars and galaxies, had to be formed within the cooling Universe.

In more detail, the model quantitatively describes the emergence of chemical elements and growth of structure from the early minute fluctuations in the matter density by gravitational collapse -- all set inside an expanding space, as follows from Einstein's general theory of gravity. Note that "expanding" means in this context that the average density of stuff and the average temperature decreases, and not necessarily that the volume increases (an infinite Universe, if it is one, cannot increase in size or volume); likewise, the Big Bang did not happen somewhere but everywhere simultaneously. For this model to work, the largest fraction in the matter soup has to be so-called dark matter. In contrast, ordinary visible baryonic matter constitutes only a small fraction but yet provides the building-blocks of stars or planets. The existence of dark matter is inferred from the effect it has on the propagation of light or on the motion and distribution of visible matter.

A direct detection of dark matter, however, is considered challenging as it appears to be only weakly interacting. Indeed, so far no strong evidence for a direct detection has been found. To physicists only weakly interacting particles, however, are nothing too unusual: the well-studied neutrinos of the standard model of physics interact only through the weak force and gravity; they cannot be "seen" since they do not interact with light in any way. Sadly, dark matter cannot be just neutrinos because neutrinos are too light and diffuse in order to explain the observable large-scale distribution of galaxies.

Distribution of millions of galaxies within two surveyed slices of today's Universe. The distribution of galaxies is thought to reflect the distribution of the dark matter. The earth is at the centre of the image. Light needs several billion years to cross the size of the survey volume. The fluctuations in the galaxy density seen here are compatible with the CMB fluctuations when we account for the gravitational collapse.

Credit: The team from the 2dF Galaxy Redshift Survey

The hot and homogeneous beginnings of the Universe are still visible today, namely through cosmological observations of the cosmic microwave background (CMB). We observe this radio radiation with basically the same spectrum in every direction of the sky (after contaminants of our own galaxy or radio point sources have been subtracted). The spectrum of this radiation is perfectly consistent with that of thermal radiation at about 3K -- a logical consequence of the Hot Big Bang that was predicted in the 1940ies. Nevertheless, it took another 30 years until the CMB was discovered by Penzias and Wilson by accident. Another robust prediction of the Hot Big Bang is the relative abundance of light elements in the early Universe which perfectly fits the observations: about 75% of the baryonic mass in the Universe is hydrogen and almost all the rest is helium. Traces of atoms that are heavier than lithium must have been baked in the cores of stars by nuclear fusion or, apparently, by colliding neutron stars, as has been found out recently.

Temparature fluctuations in the CMB sky; they reflect the early density fluctuations from a time when the Universe was about 300.000 years old.

Credit: NASA and the WMAP team

On closer inspection of the CMB, there are tiny angular fluctuations in the temperature of the order of 0.00001 degree Celcius. These reflect, among other things, the small variations in the matter density during the time the Universe became cold enough to be transparent for (non-ionising) electromagnetic radiation, which was roughly 300.000 years after the Big Bang. The LCDM successfully links these fluctuations, in particular their dependence on size, to the distribution of matter and galaxies in the Universe of today, about 13 billion years later. Additionally, it consistently links the frequency spectrum of the CMB fluctuations to the observable expansion history of the Universe.

Moreover, LCDM quantitatively describes the observed gravitational deflection of light from far distant galaxies by the intervening matter. Gravitational lensing is a relatively new brand of science that exploits this light-deflection effect in order to probe the gravitational field produced by (mostly) dark matter. Importantly, the physical nature of lensing matter is irrelevant: if it has mass, it has a gravitational attraction; and thus it is deflecting light rays. This makes gravitational lensing a unique probe for cosmologists.

Over the course of the last ten years or so, gravitational lensing has come of age, and we have gained a flavour of what can be achieved by studying this effect. To name a few applications, we have directly mapped the dark matter distribution; studied the matter density profiles of galaxy clusters and determined their mass; measured the mismatch between the galaxy and matter distribution for various galaxy populations; measured the mean cosmic matter-density and the amplitude of its fluctuations with time; and very recently confirmed the accelerated expansion of the cosmos that is also seen in studies of the CMB, as already mentioned, but also in studies of supernova SNIa or the baryonic accoustic oscillations in the galaxy distribution.

Sketch of light rays travelling through an inhomogeneous
Universe. The net effect of the small continuous deflections by local
matter over- or under-densities is that the original image of a galaxy
(left) gets distorted (right). The cumulative deflections are hence
encoded inside galaxy images.

Credit: S. Colombi (IAP) and the CFHT team

An important discovery with gravitational lensing has been made for the famous Bullet Cluster. This cluster is the remnant of the collision of two large galaxy clusters. In this accident, the collision-less content of the clusters, still attached to the visible two clusters of galaxies (left and right in the figure below), was displaced from the baryonic content, which is visible by the X-ray emission of hot gas between the clusters of galaxies (depicted as red). The Bullet Cluster therefore provides strong evidence for the picture that cluster galaxies are gravitationally glued together by a collision-less form of matter. To explain, everything that is not collision-less, such as ordinary baryonic gas, has to end up between the clusters because of the interactions between the gas atoms and molecules. All collision-less matter, on the other hand, resides with the (essentially collision-less) galaxies. Now, the analysis of the lensing signal shows us that almost all mass in this system is still with the galaxies after the collision (depicted as blue), and the mass of the gas (red) is negligible in comparison. Therefore, the matter that glues galaxies together in a cluster has to be collision-less, which are exactly the attributes that we have to postulate for dark matter.

Optical image of the bullet cluster overlaid with a mass density map (blue), inferred from lensing, and a X-ray intensity map (red), originating from the gas component that was shock-heated by the collision of the colliding clusters. After the collision of two galaxy clusters (left and right in blue), the original intra-cluster gas was stripped off and deposed between the clusters.

Credit: M. Markevitch et al. (X-ray) & D. Clowe et al. (optical; lensing map)

Gravitational lensing is also a great tool to study galaxy physics. Galaxies are thought to have formed inside the wells of the dark matter gravitational potential: The wells suck up the baryonic matter, cool it (if the cooling rate allows), and form stars. We call the dark matter concentrations that provide these wells "dark matter haloes". During the evolution of the matter density fields, these haloes and their galaxy content merge to form larger and larger haloes. While the evolution of dark matter halos with time is well understood within the framework of LCDM, the formation/evolution of galaxies is notoriously difficult to predict due to the complex baryonic physics involved. The next crucial test of LCDM is whether it is capable of explaining the large variety of galaxy properties.

To make progress here, theorists will have to refine and improve their models of galaxy physics in the next years to come, as much as observers will have to gather more data on the galaxy-dark matter connection.

Surely gravitational lensing will be an important player in this endeavour.