The work I have performed for this thesis covers a number of different areas of astronomy. My work on the LSBs is an in-depth exploration of the gas morphology and kinematics of these unusual objects. The focus is on the star forming abilities of the galaxy sample, but the HI observations also reveal a rich variety of gas morphology and motion, which are not necessarily apparent at other wavelengths. As well as being LSB, they are neutral gas-rich objects. As such they are more suited to being detected by HI surveys than by optical surveys. The Arecibo Galaxy Environment Survey (AGES) is a survey that expects to detect a large number of this type of galaxy and a large proportion of the thesis has been dedicated to my involvement with the survey. This has included designing and implementing the observing strategy, testing the data reduction pipeline, producing the final data, testing their quality and examining the efficiency of the survey detection methods. I have demonstrated some uses the AGES data can be put to from discovering hitherto undetected galaxies, through measuring the cosmic distribution of neutral hydrogen, to evaluating the cosmic significance of dark galaxies. The last point forms the central theme throughout the thesis. Of the 69 detections made by AGES in the first two datacubes, 55 do not have previous Hi measurements and 26 are previously uncatalogued. An Hi mass function (HIMF) was produced from the first galaxies detected by AGES. The best fit line to the data was a Schechter function with the parameters: a = -1.28 0.17, = 0.0076 0.0027 and M* = 7.7 1.8 x 109 M0. Using the HIMF the calculated value for the space density of neutral hydrogen was found to be pni = 7.5q3 x 107MQ Mpc-3, which is consistent with previous measurements. The overall contribution of HI to the Universal energy density was calculated from this value and found to be Qhi = 4.1 J x 10-4, confirming previous measurements of this value. Assuming the two candidates are indeed optically dark, based on this value the contribution of dark galaxies was calculated to be 4% to their respective mass bins, and less than 0.5% overall. The number density was found to be ndark 6.6 x 10-4 Mpc-3. Hence it would appear that dark galaxies are not numerous enough nor of substantial mass to be able to account for the 'missing mass' or to be able to fully represent the high number of dark matter haloes produced by CDM simulations.

QUaD is a ground based bolometric polarimeter which observed the cosmic microwave background (CMB) at 100 and 150 GHz for three austral winters. Two different techniques field differencing and ground template subtraction, were used to analyze the data providing highly accurate maps and power spectra of both temperature and polarization. These measurements were then used to constrain cosmological parameters and in conjunction with the results of several other experiments further our knowledge and understanding of cosmology in the early universe. This thesis presents an investigation into the astrophysical phenomena in the fore ground of QUaD's observation region. Although an area with minimal foreground was selected for the experiment, nevertheless between QUaD and the CMB there are various sources of millimeter radiation that could potentially contaminate the measurements. Template temperature maps of the dust and synchrotron radiation in the region are extrapolated to CMB frequencies and correlated with QUaD's maps to quantify the extent of any structure similarity detected. These are compared with the correlation expected by chance in a single universe to confirm the cleanliness of the field and highlight the robustness of the published CMB results. The residual signal persistently observed in the difference spectrum between QUaD's 100 GHz and 150 GHz frequency data is fitted against template spectra for the far infrared background (FIRB) and the Sunyaev-Zel'dovich (SZ) effect. The fitting analysis serves to constrain the parameters, b (the galaxy clustering bias), and a8 (the rms mass fluctuations in a Sh l Mpc sphere) upon which the amplitudes of the template spectra are dependent. The contribution to the power from the predicted background of radio point sources is subtracted before a simultaneous fit to the two parameters is carried out. It is found that cr8 = l.lllo!o6 = 3-2T to 23. Whilst these values are shown to be consistent with other recent astronomical results, the amplitudes of the FIRB and SZ effect they imply are, once extrapolated to high-, incompatible with the QUaD data on these angular scales. This suggests that some other factors perhaps instrumental systematics, features in the analysis pipeline, or indeed other astrophysical phenomena could also be responsible for the signal seen in the frequency difference spectrum.

No description available.

The 6.67 GHz methanol maser is uniquely associated with the phenomena of the earliest stages of high mass star formation, high mass protostellar objects (HMPOs), hot molecular starless cores (HMSCs), and compact HII regions. An unbiased survey for these masers in the plane of the Galaxy, the Methanol Maser Multibeam (MMB) Survey, is outlined. It is noted that for the subsequent exploration of these regions highly spatially resolved mid and far infrared observations are required, since this is where the emission from HMSCs and HMPOs peak. I discuss a new far infrared camera, THUMPER, which was designed and built in Cardiff. THUMPER was capable of making such observations of HMSCs and HMPOs. The observations that were carried out with THUMPER on the JCMT are presented. These are the characterisation of the atmospheric transmission at 200 /im on Mauna Kea and the first 200 /zm ground based astronomical images. I present the results of the MMB survey and compare the maser detections with various infrared parameters. The spectral indices (a) of the infrared sources are calculated and a correlation between a and the flux of the associated maser is found. This is explained in terms of the mass of dust present and age of the source. The more dust associated with the source, the brighter the maser.

In this thesis, I present SCUBA polaximetry data of three prestellar cores, L1544, L43 and L183. I also present SCUBA scan-map observations of three molecular clouds, Orion B, L1689, and RCrA. I use the Chandrasekhar-Fermi technique to measure the magnetic field strength in LI544, L43 and LI83, and find that each core is approximately magnetically critical or slightly supercritical. I measure the core mass function for prestellar cores in the Orion B molecular cloud and find that above 1 M0, the data are best fitted by a power-law of the form diV/dlog M cx M 15. The mass function has a peak at approximately 1 Mq. This core mass function is comparable with the Salpeter slope of M-1,35. I also determine the mean surface density of companions Ec(0) for the Orion B region, and find that it is best fitted by a power law of the form £c(0) oc 0 03. This is smaller than the slope of 9 6 that is typically found for young stars in most molecular clouds. However, this slope is consistent with my data at the extreme limits of my la error bars. My value of T,c(0) oc 0 3 is comparable with the slope of 9 0 2 that is found for the nearby Orion Trapezium association. I find that the L1689 molecular cloud has a comparable rate of star formation to the neighbouring and more massive LI688. In addition, the star formation in LI689 occurs in a number of filaments that are parallel both to each other, and to similar filaments in L1688. I conclude that these observations indicate that the star formation is being influenced by the nearby Sco OB2 association. I compare the RCrA data to previous surveys of the region, and discover a previously unknown Class 0 source.

Interstellar dust affects our view of the Universe, with half of the starlight emitted since the Big Bang being absorbed by dust grains and re-radiated in the Infra-Red. This thesis is concerned with understanding the origin of interstellar dust and determining the relative importance of different dust formation sites in the Universe. The sources and maximum amount of dust in early galaxies are investigated using published extended atmosphere models, stellar evolution tracks and nucleation conditions. A chemical evolution model is modified to include the estimated condensation efficiencies. The implications are investigated and we show that a supernovae source is required to produce large amounts of dust in galaxies. The atmospheres of AGB stars evolve too slowly to be responsible for high redshift dust. If SNe are not responsible for dust, then significant dust masses can only be generated at z > 5 by galaxies with very efficient star formation and no dust destruction. The first sub-millimetre analysis of the galactic remnant Kepler is presented with 0.3 3 M0 of dust, depending on the absorption properties. This a 1000 times more than previous Infra-red observations found. The sub-mm emission anti-correlates with the other wavelengths, including the warm dust and the most likely origin is from freshly synthesised dust in the supernova and dust formed from the interaction of the supernova shock with the ambient medium. Iron needles as a possible origin of the sub-mm emission are investigated. Finally, the sub-mm observations of the massive LBV star, rj Carinae are presented to show that the winds of massive stars may also be important contributors to the dust budget. We conclude that supernovae, or their massive star progenitors, are a dominant contributor to interstellar dust and make suggestions for future work.

Current theories of large scale structure and galaxy formation predict the existence of numerous low mass dark matter haloes in the Universe today. If these haloes contain sufficient stars they should be detectable as low luminosity stellar systems or dwarf galaxies. We have searched for these objects in four regions of increasing density - the general field, the area around a giant spiral galaxy, the low density Ursa Major cluster, and the high density Virgo cluster. Using identical deep optical data covering a total of 602 and probing fainter magnitudes than has been done previously, we used identical selection and detection methods to compare the dwarf galaxy populations in these different environments. We found substantially more dwarfs per giant galaxy in the Virgo cluster (20:1) compared to the field environment (6:1 max). A comparison of the HI properties and (B-I) colours for the objects for which we had additional data also showed that in general, the cluster objects are redder and gas poor compared to the objects in the field. We discuss the possible mechanisms which may have resulted in creating a population of cluster dwarf galaxies, which would explain the high number density which we found in our data. It is likely that a combination of tidal interactions and transformation of infalling dlrrs into dEs will result in the large population of cluster dwarfs. Conclusive evidence regarding their formation must now be obtained by a more detailed investigation of their stellar populations. The lack of dwarf galaxies in the field region is likely to be due to the effect of inefficient star formation in the field environment compared to the cluster. Thus the low mass dark matter haloes predicted by CDM models must still be 'dark' and can only be identified by further deep HI studies of the field environment, and future gravitational lensing studies of substructure.

In this thesis we investigate the influence of certain physical effects on the collapse and fragmentation of isolated, low-mass, low-turbulence cores, in particular on the mass distribution, binary statistics and kinematics of the resulting stars. We perform numerical simulations using a Smoothed Particle Hydrodynamics code to model this mode of star formation. Firstly we model acoustic oscillations of a self-gravitating isentropic monatomic gas sphere using our SPH code and find that if the smoothing lengths are adjusted so as to keep the number of neighbours in the range AAu, NNE1B should be set to zero, to reduce the level of numerical dissipation and diffusion. We suggest that this should become a standard test for codes simulating star formation, since pressure waves generated by the switch from approximate isothermality to approximate adiabaticity play a crucial role in the fragmentation of collapsing cores. We perform a large ensemble of SPH simulations of cores having different levels of turbulence, using a new, more realistic treatment of thermodynamics, developed by Stamatellos et al. (2007), which takes into account the thermal history of protostellar gas and captures the thermal inertia effects. We compare the results with simulations using a standard barotropic equation of state. We find that increasing the level of turbulence generally tends to reduce the fraction of the core mass which is converted into stars, and increase the number of stars formed by a single core. Using the new treatment results in more protostellar objects being formed, and a higher proportion of brown dwarfs. Of the multiple systems that form, they tend to have shorter periods, higher eccentricities and higher mass ratios. We also note that in our simulations the process of fragmentation is often bimodal, in the following sense. The first protostar to form is usually, at the end, the most massive, i.e. the primary. However, frequently a disc-like structure subsequently forms round this primary, and then, once it has accumulated sufficient mass, quickly fragments to produce several secondaries. We believe that this delayed fragmentation of a disc-like structure is likely to be an important source of very low-mass stars in nature (both low-mass hydrogen-burning stars and brown dwarf stars). We also model the evolution of an ensemble of prestellar cores in the Ophiuchus Main Cloud using initial conditions for the sizes and levels of turbulence constrained by the observations of Motte et al. (1998) and Andre' et al. (2007), and the recently revised core masses of Stamatellos et al. (2007). We find that star formation in these core is extremely efficient with typically the formation of a single star, but we also see the formation of multiple systems in a number of cores. We find that the number of stars formed by a core is highest if the core has high mass, and/or if it has a high initial level of turbulence, and/or if it starts from a low initial density. We explain why. Finally we explore the effect metallicity has on the mass distribution and binary statistics of stars formed from low-mass low-turbulence cores. We find that reducing the metallicity decreases the number of stars formed from a single core and reduces the number of brown dwarfs formed. It also reduces the binary frequency.

Using an N-body/Tree-code and Smoothed Particle Hydrodynamics, the influence of ram pressure stripping and harassment on an infalling star-forming galaxy population is investigated. In combination these mechanisms act to strip late-type dwarfs of their gas content in less than 0.5 Gyrs, causing a cessation of star-formation. As a result, late-type dwarfs evolve to colours typical of cluster dwarf ellipticals in less than 2 Gyrs, although the period in which they would be classified as transition objects lasts less than 1 Gyr. The rapid removal of a significant fraction of the dwarfs disk mass by ram pressure stripping causes morphological transformation of the stellar component, producing rotation-to-dispersion ratios in agreement with observations of cluster dwarf ellipticals. A Monte-Carlo simulation is utilised to quantify the effects of harassment, and strong, disruptive tidal encounters are found to be rare. The typical observable con sequences of harassment are found to be minor, producing very low surface brightness features (visible at surface brightnesses limits of > 30/j,b arcsec 2), and a reduction in dynamical mass-to-light ratios by roughly a factor of 2. The influence of harassment is found to be highly sensitive to the disk scale-length in exponential disks, and much less so to the mass of the galaxy. In general, ram pressure stripping is found to dominate the environmental influences on infalling late-type dwarfs, while the inclusion of the harassment model produces second-order effects only. Ram pressure stripping appears capable of forming the observed cluster dwarf population, both in colour and in morphology.

This thesis concerns the use, in gravitational wave data analysis, of higher order wave form models of the gravitational radiation emitted by compact binary coalescences. We begin with an introductory chapter that includes an overview of the theory of general relativity, gravitational radiation and ground-based interferometric gravitational wave detectors. We then discuss, in Chapter 2, the gravitational waves emitted by compact binary coalescences, with an explanation of higher order waveforms and how they differ from leading order waveforms we also introduce the post-Newtonian formalism. In Chapter 3 the method and results of a gravitational wave search for low mass compact binary coalescences using a subset of LIGO's 5th science run data are presented and in the subsequent chapter we examine how one could use higher order waveforms in such analyses. We follow the development of a new search algorithm that incorporates higher order waveforms with promising results for detection efficiency and parameter estimation. In Chapter 5, a new method of windowing time-domain waveforms that offers benefit to gravitational wave searches is presented. The final chapter covers the development of a game designed as an outreach project to raise public awareness and understanding of the search for gravitational waves.