No description available.

No description available.

This thesis studies the origin of local S0 galaxies and their possible links to other morphological types. To address these issues, two different approaches have been adopted: a detailed study of the stellar populations of S0s in the Fornax Cluster and a study of the Tully-Fisher Relation (TFR) of local S0s in different environments. The analysis of the central absorption line indices of 9 S0 galaxies in the Fornax Cluster indicates that they correlate with central velocity dispersions (sigma0). However, the stellar population properties of these S0s indicates that the observed trends seem to be produced by relative differences in age and alpha-element abundances and not in metallicity as previous studies have found in ellipticals. The observed scatter in the line indices versus sigma0 relations can be partially explained by the rotationally-supported nature of many of these systems. It was also confirmed that the dynamical mass is the driving physical property of all these correlations and in our Fornax S0s it has to be estimated assuming rotational support. A study of the local B- and Ks-band TFR in S0 galaxies shows that these objects lie systematically below the TFR for nearby spirals in both the optical and infrared bands. This offset can be crudely interpreted as arising from the luminosity evolution of spiral galaxies that have faded since ceasing star formation. However, a large scatter is also found in the S0 TFR, which means that these galaxies cannot have formed exclusively by this simple fading mechanism after all transforming at a single epoch. For the Fornax Cluster data, the offset from the TFR correlates with the estimated age of the stars in the centre of individual galaxies implying that part of the scatter in the S0 TFR arises from the different times at which galaxies began their transformation.

One of the most fundamental observational probes of galaxy evolution is determining the build-up of stellar mass. However, around half of all energy ever emitted from galaxies has been absorbed and reprocessed by dust, which is an end-product of stellar evolution. In order to obtain a more complete understanding of galaxy evolution, sensitive observations in the far-infrared and submillimetre are required where the dust emission peaks. Previous surveys have found galaxies were significantly dustier at earlier times, but the cause of this evolution, and the origin of the dust, are hotly debated topics in astrophysics. With the Herschel Space Observatory, a complete census of the dusty galaxy population has now recently been obtained. In this thesis I investigate the properties of the diverse dusty galaxy population via a panchromatic approach, utilising data from the UV to the submillimetre to study galaxy evolution. Using the first unbiased survey of dust in the local Universe, I explore the properties of galaxies in the local Universe as a function of morphology and highlight particularly interesting populations which are traditionally thought to be passive. The star-formation histories, dust content and environments of dusty early-type galaxies and passive spirals are investigated. I show that dusty early-type galaxies comprise a small minority of the general early-type galaxy population (5.5%), and harbour on average 5.5x10^7 M_sun of dust, which is comparable to that of some spiral galaxies in our sample. I compare these dusty populations to control samples to investigate how these galaxies are different to the general galaxy population. High redshift submillimetre galaxies are the most actively star-forming and dusty galaxies in the Universe. Constraining the properties of these galaxies is important for understanding the evolution of massive galaxies and galaxy evolution models in general. Using panchromatic data from the UV to the submillimetre, I explore the physical properties of a sample of ~250um rest-frame selected galaxies at high redshift, and compare them to dusty galaxies at low redshift selected in a similar way, to investigate the differences in the dusty galaxy populations over cosmic time. I find high redshift dusty galaxies have significantly higher star-formation rates and dust masses than z<0.5 dusty galaxies selected to have a similar stellar mass. Galaxies which are as highly star forming and dusty as those at z~2 are rare in the local Universe. My results support the idea that the most dusty galaxies at high redshift are a heterogeneous population, with around 60% of our sample consistent with secular evolution, and the other 40% of galaxies are starbursting, possibly merger-driven systems. The origin of dust in galaxies at both low and high redshifts presents a challenge to current theories of galaxy evolution. Recent work has revealed a `dust budget crisis', whereby the mass of dust observed in galaxies at low and high redshift cannot be accounted for by stellar mass loss from low-intermediate mass stars. I tackle this challenge using chemical evolution modelling of the high redshift submillimetre galaxies, with a detailed treatment of the star-formation histories and the dust sources and sinks in these galaxies. It is clear that a significant mass of dust must be from supernovae and/or grain growth; however, the origin of dust in high redshift dusty galaxies remains uncertain. I also consider the impact of inflows and outflows of gas, and the effect of changing the IMF on the physical properties of high redshift dusty galaxies.

This thesis describes the properties and evolution of massive (Mstellar ≥ 1011h−2 70 M⊙) galaxies at 0 < z < 3, including their relationship to lower mass systems. Present-day massive galaxies are composed mostly of early-type objects, although it is unknown whether this was also the case at higher redshifts. In a hierarchical assembling scenario the morphological content of the massive population is expected to change with time from disk-like objects in the early Universe to spheroid-like galaxies at present. We first probe this theoretical expectation by compiling a large sample of massive galaxies in the redshift interval 0<z<3. Our sample of 1082 objects is composed of 207 local galaxies selected from the Sloan Digital Sky Survey, plus 875 objects observed with the HST from the POWIR/DEEP2 Survey and the GOODS NICMOS Survey. 639 of our objects have spectroscopic redshifts. Our morphological classification is done in the V-band restframe both quantitatively (using the S´ersic index as a morphological proxy) and qualitatively (by visual inspection). Using both techniques we find a significant change in the dominant morphological class with cosmic time. The fraction of earlytype galaxies among the massive galaxy population has changed from ∼20-30% at z∼3 to∼70%at z=0. Spheroid-like galaxies have been the predominantmorphological massive class only since z∼1. This morphological evolution is so far based on the detailed morphological analysis of these objects, which ultimately rests on the shape of their surface brightness profiles. To explore the consistency of this scenario, we examine the kinematic status of a small subset of these galaxies. We have observed in the H-band 10 massive galaxies at z ∼ 1.4 with the Integral Field Spectrograph SINFONI at VLT. Our sample of galaxies have been selected purely by their photometric stellar mass without accounting for any morphological criteria a priori, and having [OII] line equivalent widths of > 15°A Abstract 5 to secure their kinematical measurements. Through a 3D kinematical spectroscopy analysis we conclude that half (i.e. 50±7%) of our galaxies are compatible with being rotationally supported disks in agreement with our previous photometric expectations. This is around a factor of two higher than what is observed in the present Universe for objects of the same stellar mass. Strikingly, the majority of these massive galaxies show clear and fairly large rotational velocity maps, implying that massive galaxies acquire rapidly rotational support and hence gravitational equilibrium. In addition, we have evidence, based on our measured velocity dispersions and imaging, to favour a picture in which minor (and major) mergers are the main driving force behind the evolution of this massive galaxy population. There is also cumulative evidence showing that the formation process for a number of these massive galaxies occur at even higher redshifts (z > 5) and that their morphological features are preserved when observing them in the UV restframe. Hence, we made use of the excellent capabilities of GNS to locate and study massive galaxies beyond z = 3 within our imaging and secondly determining whether the strong masssize relation found for the most massive objects holds as well for lower mass objects. Our findings show the extreme compactness of massive objects at z > 3 and only a moderate evolution in size below our 1011M⊙ mass limit.

Studies of the history of the Universe are for a large part concerned with mapping the evolution of galaxies over cosmic time. Beginning from the seeds of density perturbations in the early Universe, and building up through gravitational and astrophysical interactions to form the wide diversity seen in the present day, galaxies allow us to observe the distribution of luminous (and dark) matter over a wide range of look-back times. A key process in galaxy evolution is the formation of stars, an activity which is readily observed by indirect means, although the detailed mechanism is not fully understood. One of the most successful methods for tracing star formation is to observe the emission from dust in galaxies. These tiny particles of carbon- and silicon-based solids resemble smoke, pervading the interstellar medium in many (if not all) galaxies, and blocking the short-wavelength radiation from hot, newly-formed stars. They re-radiate this energy as far-infrared radiation (wavelengths ~10-1000 microns), which can be detected from sources throughout the Universe by telescopes such as the Spitzer and Herschel space observatories. The spectral form of this radiation varies from one galaxy to another, depending on many factors such as the activity within the galaxy, the amount of dust, and the sources heating the dust. Hence, with careful interpretation, we can use these observations to trace the star-forming activity and dust mass in different types of galaxies from early times through to the present day. In this thesis I describe three projects, each of which utilises multi-wavelength datasets from large surveys to probe the dust emission from samples of galaxies at different cosmic epochs, and explore the relationship between dust emission and other galaxy properties. The first project samples the most massive galaxies at a range of redshifts spanning the peak era of star formation, and investigates the correlation between far-infrared and radio emission. I use a `stacking' methodology to avoid bias towards the brightest star-forming galaxies, and show that the far-infrared and radio tracers of star formation agree up to high redshifts in typical massive galaxies. In the second project I apply the stacking method to a large sample of low-redshift galaxies selected from a major optical survey spanning the last four billion years of evolution. I make use of the largest ever sub-millimetre imaging survey to produce a detailed and unbiased census of the dust mass in ordinary galaxies as a function of optical brightness, colour and look-back time. I show that the luminosity and temperature of dust is a strong function of galaxy mass and colour, while the dust masses of all galaxy types have decreased rapidly over the time span probed. The final project focuses on a small sample of nearby galaxies and utilises data obtained and reduced by myself to probe the molecular-gas content of galaxies selected to have large dust masses. This study addresses questions about how well the cold dust, traced by the sub-millimetre wavebands of Herschel, is correlated with the cold gas, which provides the fuel for ongoing star formation. The thesis demonstrates the utility of statistical techniques for large surveys, and also contains aspects of data reduction and extensive discussion of the astrophysical interpretation of results. Through these various analyses I show that dust emission can provide a valuable window on the growth of galaxies through star formation. The work contained herein represents significant progress in the field of observational extragalactic astronomy, including work recently published in the scientific literature in two collaborative research papers led by myself, in addition to a third paper that I am currently preparing.

In this thesis, we explore the effect of the galaxy environment on the physical size and structure of the stellar distribution for relatively local galaxies (z < 0.3) using Hubble Space Telescope/Advanced Camera for Surveys imaging and data from the Space Telescope A901/2 galaxy evolution survey (STAGES). We determine the effect of the environment on the size of the stellar distribution (i.e. galaxy sizes) by comparing the stellar-mass-size relations in the field and cluster environments for different Hubble-type morphologies. For elliptical, lenticular, and high-mass (M* > 10^10 M_sun) spirals, we find no evidence to suggest that a galaxy's size (i.e. effective radius a_e) is dependent on the environment. This result suggests that internal drivers are responsible for any potential size evolution inherent to these galaxies. However, for intermediate-/low-mass spirals (M* < 10^10 M_sun) we do find some evidence for a possible environmental effect, with the mean galaxy size (<a_e>) being ~15-20 per cent larger in the field than in the cluster. This result is driven by a population of low-mass, large-a_e field spirals (observed to contain extended stellar discs) that are largely absent from the cluster environments. This difference implies that the fragile extended stellar discs of these spiral galaxies may not survive the environmental conditions in the cluster. We expand on this result by investigating the effect of the environment on the structure of galactic discs in spiral and S0 galaxies. Using V-band radial surface brightness mu(r) profiles, we identify break features in the stellar disc (down-bending break - truncation; up-bending break - antitruncation) and evaluate their dependence on the galaxy environment. For both spiral and S0 galaxies, we find no evidence to suggest an environmental dependence on the frequency of these break features. We also find no evidence to suggest an environmental dependence on the scalelength h of pure exponential discs, or the break strength T (outer-to-inner scalelength ratio) of broken exponential discs. These results indicate that the stellar distribution in the outer regions of spiral/S0 galaxies is not significantly influenced by the galaxy environment. In our structural analyses, one interesting observation was that truncated mu(r) profiles (down-bending breaks) are very rare in S0s; whereas in spiral galaxies they are commonplace. We expand on this result by comparing the structural properties of the disc (scalelength h, break strength T, break surface brightness mu_brk) in spiral and S0 galaxies. In these comparisons, we find no evidence to suggest that the scalelength h of pure exponential discs or the break surface brightness mu_brk of broken exponentials is dependent on the galaxy morphology. However, we do find some evidence to suggest that the break strength T is smaller (weaker) in S0s compared to spiral galaxies. This result suggests that some process inherent to the morphological transformation of spiral galaxies into S0s does affect the structure of the stellar disc causing a weakening of mu(r) breaks and may even eliminate truncations from S0 galaxies. In additional structural comparisons, we also find that the fraction of exponential bulges is the same (~20 per cent) in both spiral and S0 galaxies, suggesting that major mergers are not driving this transformation. Finally, we complement our structural analyses with an assessment of whether the excess light in the outer regions of antitruncated (up-bending) mu(r) profiles is caused by an outer exponential disc or an extended spheroidal component: we use bulge-disc decomposition in order to achieve this. For spiral galaxies, in the vast majority of cases, evidence indicates that the excess light at large radii is related to an outer shallow disc. We thus conclude that in the majority of spiral galaxies, antitruncated outer stellar discs cannot be explained by bulge light and thus remain a pure disc phenomenon. However, for S0s, bulge light can have a significant effect in the outer regions of the mu(r) profile. In approximately half of S0 antitruncations, the excess light at large radii can be entirely accounted for by light from an extended spheroidal component. These results suggest that as a galaxy evolves from a spiral into an S0, the galaxy naturally evolves into a more bulge-dominated system. We suggest a fading stellar disc (e.g. caused by gas stripping and the termination of star formation) is consistent with this result. In conclusion, our environmental studies indicate that the environment has little direct affect on the size and structure of a galaxy's stellar distribution. This result implies that physical processes directly affecting the structure of the stellar distribution (e.g. mergers or harassment), are not driving the observed morphology-density relation. With respect to both our environmental and morphological studies, we can conclude that more subtle processes acting on the gaseous component of a galaxy (e.g. ram-pressure stripping) are more likely to play an important role in the origin of the morphology-density relation and the transformation of spirals into S0s.

This thesis is concerned with the targeting of emission line galaxies with FMOS (Fibre Multi-Object Spectrograph) to determine properties of star forming galaxies at redshift ~1.5, and provide measurements of the growth rate of large-scale structure through Redshift Space Distortions (RSDs). I also consider the opportunities of targeting the passive galaxy population at high redshift, through measurements of their continuum. I start with the extensive broad-band photometric data available in the UKIDSS-UDS (United Kingdom Infrared Telescope Deep Sky Survey - Ultra-Deep Survey) field which is used to produce a band-merged catalogue, later used for determining photometric redshifts. In producing this catalogue, I approach the issue of source confusion present in the deep Spitzer imaging using z-band priors on profile position and shape and an iterative Expectation-Maximisation algorithm. Photometric redshift estimates are compared against colour selections as potential targeting techniques for a wide-area redshift survey with FMOS. Different photometry survey areas are considered, and the quality of selection given the available broad-band data tested, by adjusting the photometric catalogue produced for the UDS. The results indicate that the SWIRE (Spitzer Wide area InfraRed Extragalactic Survey) fields are too small to provide adequate sources with a consistent selection mechanism. The CFHTLS (Canada-Frace-Hawaii Telescope Legacy Survey) would have a large enough area given deeper z'-band imaging, and SWIRE-depth coverage in the Spitzer 3.6μm and 4.5μm bands. I present FMOS commissioning data obtained for the UDS field, including the spectroscopic targeting of sources form the High-Z Emission Line Survey (HiZELS). With this data, I am able to test the current quality of flux calibration using cool stars targeted simultaneously and the level of systematic errors left by sky-subtraction. The sample of HiZELS sources selected to place Hα at z~1.45 show low contamination from other emission lines, and only one out of 9 targets assigned a redshift has any indication of AGN activity. Finally, I present longslit observations of faint, passive galaxies at redshift z~1.9, selected as members of a possible cluster, JKCS 041, selected from broad band colours. One object was observed with high enough signal to noise to constrain the position of the 4000 Å / Balmer break, providing a tighter constraint on the photometric redshift of 1.8867 ^+0.0034 -0.0117.

Superclusters appear as large-scale structures in the form of a network of filaments, and can be up to 100h\(^{-1}_{100}\) Mpc in extent. In this dissertation, we investigate in detail the spatial structure of the three richest superclusters of galaxies closer to us then z=0.1. We investigate the rate of star formation in galaxies at various positions among the filaments and clusters in the Pisces-Cetus Supercluster. We use an index of star formation derived from a principal component analysis of optical spectral. We have shown that galaxies which are members of these filaments, show a steady decline in star formation rate, from the periphery of a cluster, into the cluster core. However, on top of this trend, we find a nearby instantaneous enhancement of the rate of star formation at ~3h\(^{-1}_{70}\) Mpc from its centre. We conclude that the most likely reason for this sudden enhancement in star formation rate is galaxy-galaxy harassment. Further work shows that the enhancement in star formation occurs mainly in the in falling dwarf galaxies (-20 < MB < -17.5) and that there is little evidence that the tidal effect of the dark matter haloes of the clusters is responsible for the enhanced star formation. The results of an analysis performed on a larger ensemble of 52 filaments were consistent with those from our smaller sample drawn from the Pisces-Cetus supercluster. We conclude this study with the analysis of a sample of spectra from the 6dF redshift survey. In the absence of spectrophotometric calibration, for these galaxies we were only able to obtain an uncalibrated star formation rate, but we could examine the effect of correction due to dust extinction, and could separate the starforming galaxies from the active galactic nuclei. From our small sample, there was interesting evidence of enhanced star formation in galaxies at similar distances from the centres of the clusters in the Shapley Supercluster.

This thesis studies a new method of constraining the mass distributions of elliptical galaxies and clusters of galaxies: gravitational redshift. The aim was to determine the types of astrophysical object in which gravitational redshift can be most readily detected and to attempt to observe the effect for the first time in a single object. Longslit stellar kinematics were combined with planetary nebulae kinematics to study the mass distribution of M87. Jeans modelling showed that, although the best-fit model gave too little mass (Upsilon = 5.34+/-0.34, beta = 0.71+/-0.03, M_halo = 2.64+/-0.92 x 10^12 M_sun), by adjusting the orbital anisotropy it was possible to construct a model that was consistent with both the kinematics and existing X-ray gas measurements. Longslit kinematics from the literature were used to attempt to determine the mass-to-light ratio of a sample of elliptical galaxies using gravitational redshift. Models were developed to calculate the expected gravitational redshift from the surface brightness profile. The best-fit mass-to-light ratios were found using this model and also obtained independently using the Jeans equation. The results were not statistically inconsistent with the expected H-band mass-to-light ratios predicted by stellar population models - the Jeans modelling gave a mean mass-to-light ratio of Upsilon_sigma = 1.67+/-0.10, while the gravitational redshift predicted Upsilon_v = 4.84+/-2.67. Integral field spectroscopy of the centre of M60 was undertaken in an attempt to detect gravitational redshift in the centre of an individual galaxy for the first time. The velocity field was summed around the galaxy's isophotes to remove the effect of rotation. Models were constructed to predict the gravitational redshift, which were then compared to the data. It was found that the shallow slope of the light profile made it impossible to detect a gravitational redshift in this case, but that the scatter of the data points suggests that a signal of a few km/s could be detected. Consideration of the models led to a better understanding of the most suitable targets for this kind of study. An analysis of the 2dF groups catalogue was made in order to attempt to determine the strength of the gravitational redshift in clusters of galaxies. A new method was developed for measuring the signal in clusters. As part of the analysis, the density distribution of the clusters was obtained. It was found that they followed an exponential profile, which scaled linearly with the size of the cluster. The gravitational redshift was used to attempt to constrain the mean cluster mass, but it was found that the errors were too large to rule out all but the largest masses with any certainty. Future studies would require either a much larger sample, or one which concentrates specifically on the most uniform, high mass clusters. Gravitational redshift offers a new approach to studying the mass distributions of galaxies and clusters that requires many fewer assumptions regarding the underlying physics than many of the current methods. Unfortunately, it also suffers from a number of potential setbacks. Recent advances in instrument technology, combined with the careful selection of suitable targets should allow gravitational redshift to become a viable tool for studying the nature and distribution of dark matter.