nterferometric observations at millimeter wavelengths provide a precious, detailed view of certain astrophysical objects. This thesis is composed of studies that both rely on and enable this technique to study the structure of planet-forming disks and soon image the closest regions around super-massive black holes. Young stars form out of a cloud of gas and dust that, before its eventual dissipation, flattens to a disk. However the disk population is diverse and recent high-resolution images have revealed a wide variety of interesting features. To understand these observations we use detailed radiative transfer models to motivate various physical scenarios. First we identify a set of traits in the disk around V4046 Sgr that marks the coupled progression of the gas and dust distributions in the presence of at least one embedded companion. Next, we investigate how the vertical temperature structure of a disk can be spatially resolved and apply our framework to observations of the disk around HD163296. Lastly, we show how large-scale radial flows of gas may be observable and question how this phenomenon might be distinguished from other scenarios such as warps or outflows. The last chapter summarizes the APHIDS project which changes the sampling rate of data taken at the SMA so that it may be used for VLBI campaigns. / Astronomy

I present a three-dimensional map of interstellar dust reddening, covering three-quarters of the sky out to a distance of several kiloparsecs, based on Pan-STARRS 1 and 2MASS photometry. The map reveals a wealth of detailed structure, from filaments to large cloud complexes. The map has a hybrid angular resolution, with most of the map at an angular resolution of 3.4′ to 13.7′ , and a maximum distance resolution of ∼25%. The three-dimensional distribution of dust is determined in a fully probabilistic framework, yielding the uncertainty in the reddening distribution along each line of sight, as well as stellar distances, reddenings and classifications for 800 million stars detected by Pan-STARRS 1. The method developed here compares observed stellar photometry with empirical stellar templates, incorporating prior knowledge about the structure of the Galaxy. I validate the per-star reddening estimates by comparison with reddening estimates for stars with both SDSS photometry and SEGUE spectral classifications, finding per-star agreement to within ∼0.15 mag out to a stellar E(B−V) of 1 mag. I demonstrate the consistency of the resulting reddening estimates with those of two-dimensional emission-based maps of dust reddening. In particular, I find agreement with the Planck τ353 GHz-based reddening map to within 0.05 mag in E(B−V) to a depth of 0.5 mag, and explore systematics at reddenings less than E(B−V) ≈ 0.08 mag. I compare the 3D map developed here to two existing three-dimensional dust maps, by Marshall et al. (2006) and Lallement et al. (2013), exploring the strengths and weaknesses of the different 3D mapping methods. The map presented here has better angular resolution than both 3D maps compared, and it has better distance resolution than Marshall et al. (2006) within ∼3 kpc, but shows radial “finger-of-God” features not contained in Lallement et al. (2013). The map can be queried or downloaded at http://argonaut.skymaps.info. I expect the three-dimensional reddening map presented here to find a wide range of uses, among them correcting for reddening and extinction for objects embedded in the plane of the Galaxy, studies of Galactic structure, calibration of future emission-based dust maps and determining distances to objects of known reddening. The method we present is not limited to the passbands of the Pan-STARRS 1 and 2MASS surveys, but may be extended to incorporate photometry from other optical and near-infrared surveys, such as WISE, Spitzer GLIMPSE, UKIDSS, SDSS (where available), and in the future, LSST and Gaia. The method can also be naturally extended to stellar kinematic data, such as that soon to be released by Gaia. / Astronomy

Accurate knowledge of the fundamental properties of stars--mass, temperature, and luminosity--is key to our understanding of stellar evolution. In particular, empirical measurements of stellar mass are difficult to make and are generally limited to stars that dynamically interact with a companion (e.g., eclipsing or astrometric binaries), a precious but ultimately small sample. We developed a technique that uses the rotation of the protoplanetary disk--a consequence of the star formation process still present around many pre-main sequence stars--to measure the stellar mass. To establish the absolute accuracy of this technique, in ALMA Cycle 1/2 we observed the few circumbinary disks around double-lined spectroscopic binary stars, enabling an independent confirmation of the total stellar mass. This comparison with radial-velocity results demonstrates that the disk-based dynamical mass technique can reliably achieve precise measurements of stellar mass on the order of 2-5\%, clearing the way for widespread application of this technique to measure the masses of \emph{single} stars. We discuss our calibration in the context of two sources, AK~Sco and DQ~Tau. Second, we developed novel statistical techniques for spectroscopic inference. Young stars exhibit rich and variable spectra; although interesting phenomena in their own right, accretion veiling and star spots complicate the retrieval of accurate photospheric properties. The subtraction of an imperfect model from a continuously sampled spectrum introduces covariance between adjacent datapoints (pixels) into the residual spectrum. For the high signal-to-noise data with large spectral range that is commonly employed in stellar astrophysics, that covariant structure can lead to dramatically underestimated parameter uncertainties (and, in some cases, biases). We construct a likelihood function that accounts for the structure of the covariance matrix, utilizing the machinery of Gaussian process kernels. This framework specifically addresses the common problem of mismatches in model spectral line strengths (with respect to data) due to intrinsic model imperfections (e.g., in the atomic/molecular databases or opacity prescriptions) by developing a novel local covariance kernel formalism that identifies and self-consistently downweights pathological spectral line ``outliers." We demonstrate some salient features of the framework by fitting the high resolution $V$-band spectrum of WASP-14, an F5 dwarf with a transiting exoplanet, and the moderate resolution $K$-band spectrum of Gliese~51, an M5 field dwarf. Direct spectroscopic inference provides one means to avoid the systematic error that results from the uncertain spectral type--effective temperature scale for low mass pre-main sequence stars when placing a star on the Hertzsprung Russell diagram. Lastly, we discuss recent progress in measuring the masses of a large sample of single pre-main sequence stars observed with the Submillimeter Array, which will double the number of disk-based dynamical mass estimates of pre-main sequence stars. With ALMA, the disk-based technique holds enormous promise to become the primary means of stellar mass for statistically large samples of pre-main sequence stars, ushering in a new era of high precision in star and planet formation studies. / Astronomy

Data from the NASA Kepler spacecraft indicate that small planets are common around the smallest main sequence stars (M dwarfs). Rocky planets transiting M dwarfs will be the best targets for atmospheric characterization with the next generation of scientific instruments. Ground based transit and radial velocity surveys, and the upcoming NASA TESS mission are expected to reveal the transiting terrestrial exoplanets that are nearest to the Sun. Understanding these worlds requires that we first understand their host stars. In this thesis, I present better estimates of the distances, masses, radii, and metallicities of these target stars. I used data from the MEarth-North transit survey to obtain trigonometric distances to 1507 mid-to-late M dwarfs with a precision of 5 milliarcseconds. I use these distance measurements to obtain better estimates of the masses and radii of these stars than available from photometry alone, and prioritize targets to monitor at high cadence for transiting planets. I find that the M dwarf census in the northern hemisphere is mostly complete to a distance of 25 parsecs for stars of spectral type M5.5V and earlier, and mostly complete for stars earlier than M7.0V out to 20 parsecs. I present calibrated MEarth optical photometry of 1844 MEarth-North targets with a typical precision of 1.5%. By combining these measurements with trigonometric distances, spectroscopic metallicities, and extant near infrared (NIR) magnitudes, I derive a color-magnitude-metallicity relation with a precision of 0.1 dex. I find that the median metallicity for Solar Neighborhood M dwarfs is [Fe/H] = -0.030 +/- 0.008, indistinguishable from the solar neighborhood G dwarfs. I present the MEarth-South discovery of LP 661-13, a low-mass double lined eclipsing binary system with an orbital period of 4.7 days. I determine the component masses to be 0.3050 +/- 0.0056 and 0.1937 +/- 0.0027 M_sun and the radii to be 0.3192 +/- 0.0037 and 0.2159 +/- 0.0061 R_sun. While each component is marginally consistent with stellar models, the sum of the radii is well constrained and is inflated 5% compared to stellar models, which cannot be ascribed to metallicity or age effects. LP 661-13 joins the small sample of low-mass stars with precisely measured masses and radii that serves as a robust test of models of fully convective dwarf stars. I present calibrated griz photometry of 150 of the MEarth-North target stars with a typical precision of 1%. I find that no combination of griz filters alone can reliably determine the metallicity of an M dwarf. However, interpolation in the (g-i, i-K) color plane can estimate the metallicity with a typical standard deviation of 0.1 dex. This precision is comparable to that from NIR spectroscopic methods that have been recently developed. Overlap between the upcoming Large Synoptic Survey Telescope (LSST) sources and the VISTA Hemisphere Survey sources will provide a sample of M dwarfs with estimated metallicities that can in turn be used as a chemical probe of the Milky Way Galaxy. / Astronomy

Multi-wavelength observations of supernovae (SN) not only probe the explosion mechanism, but also carry information about the configuration of the star at the moment of collapse and the mass-loss history of the progenitor system in the years immediately preceding its death. The study of supernovae therefore offers a rare observational view of the final stages of stellar evolution. As a result, the recent advent of wide-field SN searches---which are discovering new classes of astronomical transients at an ever-increasing rate---has both expanded the types of stellar systems that we can directly probe and challenged theoretical models of the late-stages of stellar evolution. This thesis explores several new regimes of transient space that have been opened by these modern, wide-field, time domain surveys. We present a series of observational studies which constrain the explosion properties, progenitor systems, and intrinsic rates for several classes of peculiar astronomical transients. First, we investigate the properties of transients that reach SN luminosities but evolve on rapid timescales. We present a detailed study of the rapidly-declining hydrogen-poor SN 2005ek, showing that it could be produced by the core-collapse of a stripped massive star with a very small ejecta mass. We then describe results from the first systematic search for rapidly-evolving and luminous transients in a wide-field survey with a rapid cadence. Using data from the PanSTARRS1 Medium-Deep Survey (PS1-MDS), we both identified a new class of rapidly-evolving transients and calculated their intrinsic rates---demonstrating that these explosions are not intrinsically rare. Second, we utilize very early UV/spectroscopic observations and detailed multi-wavelength follow-up of the stripped-envelope core-collapse SN 2013ge to place constraints on the final configuration of the progenitor system. Finally, we investigate the progenitor stars and mass-loss mechanism that operates in a class of luminous SN that explode within a dense circumstellar medium (Type IIn SN). This is accomplished through a joint analysis of the explosion properties and host galaxy environments for a large number of events, including the full sample of Type IIn SN discovered by the PS1-MDS. We find that luminous Type IIn SN explode in a wide range of galaxies which are are robustly distinguished from the host galaxies of hydrogen-poor super-luminous SN in being more luminous, massive, and metal-rich. Furthermore, we show that a significant fraction of luminous Type IIn SN explode in the bright central regions of their host galaxies, indicating that the physical conditions for star formation in these environments are conducive for the formation of luminous Type IIn progenitors. / Astronomy

Despite the prevalence of M dwarfs, the smallest and most common type of main sequence star, their sizes, compositions, and ages are not well-constrained. Empirical determination of these properties is important for gaining insight into their stellar structure, magnetic field generation, and angular momentum evolution. I obtained low-resolution (R = 2000) near-infrared spectra of 447 nearby mid-to-late M dwarfs. I measured their absolute radial velocities with an accuracy of 4.4 km/s by exploiting telluric lines to establish an absolute wavelength calibration. I estimated their metallicities from the equivalent width of the sodium absorption feature at 2.2 μm to a precision of 0.12 dex, and from 2MASS colors to a precision of 0.15 dex. Using stars with radii measured from interferometry, I showed that the equivalent widths of aluminum and magnesium absorption features can be used to infer K and M dwarf temperatures to a precision of 69 K, and radii to 0.027 R⊙. I applied these relations to planet-hosting stars from Kepler, showing that the typical planet is 15% larger than inferred when adopting stellar parameters from other recent catalogs. Using photometry from the MEarth-North Observatory, I measured rotation periods from 0.1 to 140 days for 387 M dwarfs. I found a prevalence of stable spot patterns, and no correlation between period and amplitude for fully-convective stars. Using galactic kinematics as a proxy for age, I found that rapid rotators (P < 10 days) are < 2 Gyr, and that the slowest are on average 5 ± 3 Gyr old. I then showed that for early M dwarfs the typical stellar rotation period at 5 Gyr coincides with the orbital period at which habitable planets are found, and I suggest that mid-to-late M dwarfs are optimal targets around which to search for habitable-zone planets. I obtained optical spectra of 247 nearby M dwarfs, and measured the strength of the chromospheric Hα emission line. I identified a well-defined boundary in the mass–period plane that separates active and inactive M dwarfs. Hα activity is therefore a simple, accessible diagnostic for stellar rotation period, and I present a mass–period relation for inactive M dwarfs. I found a significant (p value < 10e−4) positive correlation between Hα emission strength and photometric variability amplitude, which implies that stars with stronger magnetic fields have both higher levels of chromospheric activity and larger or more abundant spots. I suggest that fully convective stars maintain rapid rotation rates and saturated magnetic activity for about 2 Gyr. They then undergo rapid angular momentum evolution upon reaching some critical threshold. Only upon reaching long rotation periods (around 70 days for a 0.2 M⊙ star) do their magnetic activity levels drop below what is required for Hα to be seen in emission. The stars I have observed in pursuit of this work are the nearest low-mass stars. As such they are the best targets around which to search for habitable, rocky worlds, and my work provides the means to constrain the sizes, temperatures, and ages of those planets. / Astronomy

This thesis addresses the distributions of baryonic matter on three scales: the outskirts of the gas and galaxy profiles in galaxy clusters, the clustering of galaxies of galaxies on large scales and its relation to the underlying matter distribution, and the extremes of the galaxy distribution: the connections between the most distant galaxies ever discovered and the closest galaxies to our own, the Local Group Dwarfs. We begin with investigations of the outskirts of galaxy clusters, where long-standing analytical models of structure formation as well as recent simulations predict the existence of steep density jumps in the gas (the 'virial shock') and dark matter profiles near the virial radius. We describe a new method for deriving models for the gas distribution in galaxy clusters, which relies on a few basic assumptions --- including the existence of the virial shock and a coincident density jump in the dark matter --- and show a resulting profile for the gas that is in good agreement both with X-ray observations of cluster interiors and simulations of the outskirts, and requires fewer parameters than the traditional three-parameter beta-model. Recent simulations have strengthened the arguments in favor of the existence of a dark matter density jump, arising from the accumulation of particles at the apocenter of their first orbit. Since cluster member galaxies are expected to follow similar collisionless dynamics as the dark matter, the galaxy density profile should show a steep density jump as well. We present evidence for a feature consistent with a density jump in galaxy density profiles constructed from photometry from the Sloan Digital Sky Survey and Hectospec (MMT) spectroscopy of cluster members and discuss avenues for probing the density jumps with future data sets. Moving to larger scales where massive galaxies of different types are expected to trace the same large-scale structure, we present a test of this prediction by measuring the clustering of red and blue galaxies at z~0.6 using the CMASS sample of galaxies from the 12th Data Release of SDSS-III. The stochasticity between these two samples is quantified via the correlation coefficient r, which can be constructed from two different statistics. Both statistics indicate that on intermediate scales (20 < R < 100 Mpc/h) there is low stochasticity between the two samples of galaxies, providing a constraint on a key systematic in using large galaxy redshift surveys for cosmology. In cosmology, dense redshift surveys permit the measurement of the scale of baryon acoustic oscillations (BAO), which appear as a modest amplification at scales of about R = 105 Mpc/h in the two-point auto-correlation function of galaxies, provided that there is a sufficiently high density of galaxies with accurately measured three-dimensional positions. As a result, due to the expense of spectroscopic observations, to date most BAO analyses have been performed at fairly low redshifts where present surveys can attain the requisite densities without sacrificing efficiency. We present a new method of measuring the BAO using the cross-correlation of a sparse spectroscopic sample with a denser, photometric sample of galaxies that will allow us to extend BAO measurements to higher redshifts than are presently accessible with spectroscopy alone. We discuss applications of this new method to current and upcoming datasets. Finally, we connect galaxies both near --- the Local Group dwarf galaxies --- and far --- the high-redshift galaxies discovered by space-based observatories like Hubble and Spitzer. We evolve the local dwarfs back in time using stellar population synthesis code and juxtapose the properties of their ancient selves against those of the galaxies already discovered at high redshift. We additionally compare the properties of the dwarfs' progenitors with the detection limits of the future James Webb Space Telescope (JWST), finding that JWST should be able to detect the progenitors of galaxies similar to a few of the brightest local galaxies. / Physics

The last few years have seen enormous progress in the field of galaxy formation and evolution. The latest generation of hydrodynamic cosmological simulations (e.g., the "Illustris" simulation) has been able to produce reasonably realistic populations of galaxies by tracking the evolution of dark matter, gas, stars, and black holes over a cosmological volume "representative" of the large-scale density field. However, such increasingly sophisticated cosmological simulations require equally sophisticated analysis tools. The first part of my thesis work consisted in developing a method for connecting galaxies across cosmic time, which results in data structures known as merger trees. My algorithm, known as SUBLINK, improves upon previous methods by making galaxies less likely to become "lost" during close interactions, and has been benchmarked in a merger tree code comparison project, with favorable results. The second part of my thesis work, and the main topic of this dissertation, consists of a series of essential and increasingly complex applications of my merger trees: (1) measuring the merger rate of galaxies, (2) finding out how galaxies acquire their stellar mass, and (3) investigating the impact of mergers on galaxy morphology. I will show how my analysis tools, in combination with the Illustris cosmological simulation, have made quantitative and statistically robust contributions to the field of galaxy formation and evolution, where galaxy mergers are known to play a fundamental role. / Astronomy

In spite of the rapid progress of numerical simulations as tools for studying large-scale structure formation, there is still a considerable degree of uncertainty regarding the details of the physical processes governing the dynamics of the baryonic matter in the Universe. In this thesis, we explore the impact that specific choices in the modeling of baryonic physics in computer simulations have on the gas content of haloes spanning more than 10 order of magnitude in mass, from the size of globular clusters with masses in the $10^4-10^7M_{\astrosun}$ range, to the most massive galaxy clusters at $10^{15}M_{\astrosun}$. We start by examining the assumptions necessary in order to consistently set up the initial conditions of cosmological simulations following the evolution of both baryons and dark matter. Using adiabatic high-resolution numerical simulations, we quantify the effect that the non-negligible relative motion of baryons with respect to dark matter at the time of recombination has on structure formation and evolution. We find that a non-zero relative velocity has a sizable impact on the number density of haloes with masses $\lesssim$ few$\times 10^7 M_{\astrosun}$ up to $z=10$, the final redshift of our simulations. Furthermore, the gas stream velocity induces a suppression of the gas fraction in haloes, which at z=10 is $\sim 10 \%$ for objects with $M\sim10^7M_{\astrosun}$, as well as a flattening of the gas density profiles in the inner regions of haloes. We further identify and study the formation of moderately long lived gas dominated structures at intermediate redshifts $10 < z < 20$, that recent analytical work has proposed as potential progenitors of globular clusters. For the remainder of the thesis, we change our focus and investigate the impact on the baryonic content of haloes of physical processes which become relevant at later stages of structure formation. Using the Illustris simulations suite, a high-resolution numerical simulation of a cosmological volume of $(106.5\,\rm{Mpc})^3$, we analyze the distribution of the ICM gas in galaxy clusters with virial masses in the $10^{13}M_{\astrosun} - 2\times10^{14}M_{\astrosun}$ range. We find a substantial spread in both structural and observable halo characteristics, particularly for the less massive objects studied, and pinpoint the source of the scatter to be the radio mode of the AGN feedback implemented. Our halo sample also indicates that the impact of AGN feedback decreases for the more massive objects, which motivates us to explore the evolution of clusters with masses above those realized in Illustris. We further pursue the investigation of the impact of the specific implementation of AGN feedback on the properties of galaxy clusters by running and analyzing the iClusters suite, a series of high resolution zoom-in simulations of 6 galaxy clusters with masses between $2\times10^{13}M_{\astrosun}$ and $2\times10^{15}M_{\astrosun}$. Their evolution is modeled by using five distinct implementations of relevant physical processes, three of which include different AGN feedback models, recently used in the literature. We confirm a decreasing trend in the impact of AGN feedback on the state and observational signatures of intracluster gas with increasing halo mass. Furthermore, we find an excellent agreement with observations for all of the models of baryonic physics implemented in the iClusters simulations, which we believe is due to the fact that observations currently have access to objects with masses above $10^{14}M_{\astrosun}$. Finally, our simulations suggest that observations of lower mass objects are needed in order to place constraints on the specific details of the implementation of AGN feedback. / Physics

The composition of planets is determined by and tightly linked to the composition of the protoplanetary disk in which they form. In the first part of my thesis, I study giant planet formation through core accretion. I show how the minimum core mass required to form a giant planet during the lifetime of the protoplanetary disk depends on the location in the disk, the equation of state of the nebular gas and dust opacity. This minimum applies when planetesimal accretion does not substantially heat the core's atmosphere. The minimum core mass decreases with semimajor axis, and may be significantly lower than the typically quoted value of 10 Earth masses, thus challenging previous studies that core accretion cannot operate in the outer disk. In the second part, I explore how the composition and evolution of protoplanetary disks may affect the formation and chemical composition of giant planets. Volatile snowlines are highly important in the planet formation process. I thus show how the snowline locations of the main carbon, oxygen and nitrogen carriers, as well as the C/N/O ratios, are affected by disk dynamics and ice morphology. Compared to a static disk, disk dynamics and ice morphology combined may change the CO and N_2 snowline locations by a factor of 7. Moreover, the gas-phase N/O ratio is highly enhanced throughout most of the disk, meaning that wide-separation giants should have an excess of nitrogen in their atmospheres which may be used to trace their origins. The large range of possible CO and N_2 snowline locations, and hence of regions with highly enhanced N/O ratios, implies that snowline observations at various stages of planet formation are crucial in order to use C/N/O ratios as beacons for planet formation zones. / Astronomy