Search

Understanding how galaxies interact with the circumgalactic medium (CGM) requires determining how galaxies morphological and stellar properties correlate with their CGM properties. We report an analysis of 66 well-imaged galaxies detected in HST and VLT MUSE observations and determined to be within $\pm$500 km s$^{-1}$ of the redshifts of strong intervening quasar absorbers at $0.2 \lesssim z \lesssim 1.4$ with H I column densities $N_{\rm H I}$ $>$ $10^{18}$ $\rm cm^{-2}$. We present the geometrical properties (S\'ersic indices, effective radii, axis ratios, and position angles) of these galaxies determined using GALFIT. Using these properties along with star formation rates (SFRs, estimated using the H$\alpha$ or [O II] luminosity) and stellar masses ($M_{*}$ estimated from spectral energy distribution fits), we examine correlations among various stellar and CGM properties. Our main findings are as follows: (1) SFR correlates well with $M_{*}$, and most absorption-selected galaxies are consistent with the star formation main sequence (SFMS) of the global population. (2) More massive absorber counterparts are more centrally concentrated and are larger in size. (3) Galaxy sizes and normalized impact parameters correlate negatively with $N_{\rm H I}$, consistent with higher $N_{\rm H I}$ absorption arising in smaller galaxies, and closer to galaxy centers. (4) Absorption and emission metallicities correlate with $M_{*}$ and sSFR, implying metal-poor absorbers arise in galaxies with low past star formation and faster current gas consumption rates. (5) SFR surface densities of absorption-selected galaxies are higher than predicted by the Kennicutt-Schmidt relation for local galaxies, suggesting a higher star formation efficiency in the absorption-selected galaxies., Comment: Accepted for publication in MNRAS, 25 pages, 19 figures

We combine datasets from the CGM$^{2}$ and CASBaH surveys to model a transition point, $R_{\rm cross}$, between circumgalactic and intergalactic media (CGM and IGM, respectively). In total, our data consist of 7244 galaxies at z < 0.5 with precisely measured spectroscopic redshifts, all having impact parameters of 0.01 - 20 comoving Mpc from 28 QSO sightlines with high-resolution UV spectra that cover H I Ly$α$. Our best-fitting model is an exclusionary two-component model that combines a 3D absorber-galaxy cross correlation function with a simple Gaussian profile at inner radii to represent the CGM. By design, this model gives rise to a determination of $R_{\rm cross}$ as a function of galaxy stellar mass, which can be interpreted as the boundary between the CGM and IGM. For galaxies with $10^8 \leq M_{\star}/M_{\odot} \leq 10^{10.5}$, we find that $R_{\rm cross}(M_{\star}) \approx 2 \pm 0.6 R_{\rm vir}$. Additionally, we find excellent agreement between $R_{\rm cross}(M_{\star})$ and the theoretically-determined splashback radius for galaxies in this mass range. Overall, our results favor models of galaxy evolution at z < 0.5 that distribute $T \approx 10^{4}$K gas to distances beyond the virial radius., Submitted to ApJ, 17 pages, 8 figures

The bimodal absorption system imaging campaign (BASIC) aims to characterize the galaxy environments of a sample of 36 HI-selected partial Lyman limit systems (pLLSs) and Lyman limit systems (LLSs) in 23 QSO fields at $z \lesssim 1$. These pLLSs/LLSs provide a unique sample of absorbers with unbiased and well-constrained metallicities, allowing us to explore the origins of metal-rich and low-metallicity circumgalactic medium (CGM) at $z 9.0$, whereas low-metallicity pLLSs/LLSs are found in more diverse locations, with one population arising in the CGM of galaxies and another more broadly distributed in overdense regions of the universe. Using absorbers not associated with galaxies, we estimate the unweighted geometric mean metallicity of the intergalactic medium to be [X/H] $\lesssim -2.1$ at $z, Key figures: 13 and 18. Modified abstract. Accepted for publication in ApJ. Updated after referee comments, expanded subsection 4.4

We present the KODIAQ-Z survey aimed to characterize the cool, photoionized gas at 2.2100) between absorbers separated by less than 500 km/s imply that the metals are poorly mixed in 14.6, Submitted to the Astrophysical Journal. Comments welcome

Galaxies evolve under the influence of gas flows between their interstellar medium and their surrounding gaseous halos known as the circumgalactic medium (CGM). The CGM is a major reservoir of galactic baryons and metals, and plays a key role in the long cycles of accretion, feedback, and recycling of gas that drive star formation. In order to fully understand the physical processes at work within galaxies, it is therefore essential to have a firm understanding of the composition, structure, kinematics, thermodynamics, and evolution of the CGM. In this white paper we outline connections between the CGM and galactic star formation histories, internal kinematics, chemical evolution, quenching, satellite evolution, dark matter halo occupation, and the reionization of the larger-scale intergalactic medium in light of the advances that will be made on these topics in the 2020s. We argue that, in the next decade, fundamental progress on all of these major issues depends critically on improved empirical characterization and theoretical understanding of the CGM. In particular, we discuss how future advances in spatially-resolved CGM observations at high spectral resolution, broader characterization of the CGM across galaxy mass and redshift, and expected breakthroughs in cosmological hydrodynamic simulations will help resolve these major problems in galaxy evolution., Astro2020 Decadal Science White Paper

The circumgalactic medium (CGM) of galaxies serves as a record of the influences of outflows and accretion that drive the evolution of galaxies. Feedback from star formation drives outflows that carry mass and metals away from galaxies to the CGM, while infall from the intergalactic medium (IGM) is thought to bring in fresh gas to fuel star formation. Such exchanges of matter between IGM-CGM-galaxies have proven critical to producing galaxy scaling relations in cosmological simulations that match observations. However, the nature of these processes, of the physics that drives outflows and accretion, and their evolution with cosmic time are not fully characterized. One approach to constraining these processes is to characterize the metal enrichment of gas around and beyond galaxies. Measurements of the metallicity distribution functions of CGM/IGM gas over cosmic time provide independent tests of cosmological simulations. We have made great progress over the last decade as direct result of a very sensitive, high-resolution space-based UV spectrograph and the rise of ground-based spectroscopic archives. We argue the next transformative leap to track CGM/IGM metals during the epoch of galaxy formation and transformation into quiescent galaxies will require 1) a larger space telescope with an even more sensitive high-resolution spectrograph covering both the far- and near-UV (1,000-3,000 \AA); and 2) ground-based archives housing science-ready data., Comment: Astro2020 Decadal Science White Paper

We characterize the metallicities and physical properties of cool, photoionized gas in a sample of 152 z, Accepted by the ApJ (September 01, 2019). The quantitative results have essentially not changed. This version includes the updates following the referee's recommendations, and in particular two new figures (Figs. 1 and 2) and some text have been added to show that the metallicity estimates of the photoionized gas are quite robust even if the gas has multiple ionized gas-phases

We present simulations from the new "Figuring Out Gas & Galaxies in Enzo" (FOGGIE) project. In contrast to most extant simulations of galaxy formation, which concentrate computational resources on galactic disks and spheroids with fluid and particle elements of fixed mass, the FOGGIE simulations focus on extreme spatial and mass resolution in the circumgalactic medium (CGM) surrounding galaxies. Using the Enzo code and a new refinement scheme, FOGGIE reaches spatial resolutions of 381 comoving $h^{-1}$ pc and resolves extremely low masses ($\lesssim 1$--$100$ Msun out to 100 comoving $h^{-1}$ kpc from the central halo. At these resolutions, cloud and filament-like structures giving rise to simulated absorption are smaller, and better resolved, than the same structures simulated with standard density-dependent refinement. Most of the simulated absorption arises in identifiable and well-resolved structures with masses $\lesssim 10^4$ Msun, well below the mass resolution of typical zoom simulations. However, integrated quantities such as mass surface density and ionic covering fractions change at only the $\lesssim 30$% level as resolution is varied. This relatively small changes in projected quantities---even when the sizes and distribution of absorbing clouds change dramatically---indicate that commonly used observables provide only weak constraints on the physical structure of the underlying gas. Comparing the simulated absorption features to the KODIAQ (Keck Observatory Database of Ionized Absorption toward Quasars) survey of $z \sim2$--$3.5$ Lyman limit systems, we show that high-resolution FOGGIE runs better resolve the internal kinematic structure of detected absorption, and better match the observed distribution of absorber properties. These results indicate that CGM resolution is key in properly testing simulations of galaxy evolution with circumgalactic observations., ApJ, in press. New appendix with results from simulation with 78pc resolution CGM at z=2.5 now included. Appendix C with additional velocity phase diagrams has been omitted to fit within the arXiv space restraints

We present the results from our COS circumgalactic medium (CGM) compendium (CCC), a survey of the CGM at z, Resubmitted to ApJ after the first referee report

We present an analysis of the metallicity distribution of the dense circumgalactic medium (CGM) of galaxies at 0.1 < z < 1.1 as probed by partial Lyman limit systems (pLLSs, 16.1 < log N(H I) < 17.2) and LLSs (17.2 < log N(H I) < 17.7 in our sample). The new H I-selected sample, drawn from our HST COS G140L snapshot survey of 61 QSOs, has 20 pLLSs and 10 LLSs. Combined with our previous survey, we have a total of 44 pLLSs and 11 LLSs. We find that the metallicity distribution of the pLLSs is bimodal at z < 1, with a minimum at [X/H] = -1. The low-metallicity peak comprises (57 +/- 8)% of the pLLSs and is centered at [X/H] ~ -1.87 (1.3% solar metallicity), while the high-metallicity peak is centered at [X/H] ~ -0.32 (48% solar metallicity). Although the sample of LLSs is still small, there is some evidence that the metallicity distributions of the LLSs and pLLSs are different, with a far lower fraction of very metal-poor ([X/H] < -1.4) LLSs than pLLSs. The fraction of LLSs with [X/H] < -1 is similar to that found in pLLSs (~56%). However, higher H I column density absorbers (log N(H I) > 19.0) show a much lower fraction of metal-poor gas; therefore, the metallicity distribution of gas in and around galaxies depends sensitively on N(H I) at z < 1. We interpret the high-metallicity ([X/H] > -1) pLLSs and LLSs as arising in outflows, recycling winds, and tidally-stripped gas around galaxies. The low-metallicity pLLSs and LLSs imply that the CGM of z < 1 galaxies is also host to a substantial mass of cool, dense, low-metallicity gas that may ultimately accrete onto the galaxies., 29 pages, 14 figures, 7 tables. Accepted for publication in the ApJ

The timing and duration of the reionization epoch is crucial to the emergence and evolution of structure in the universe. The relative roles that star-forming galaxies, active galactic nuclei and quasars play in contributing to the metagalactic ionizing background across cosmic time remains uncertain. Deep quasar counts provide insights into their role, but the potentially crucial contribution from star-formation is highly uncertain due to our poor understanding of the processes that allow ionizing radiation to escape into the intergalactic medium (IGM). The fraction of ionizing photons that escape from star-forming galaxies is a fundamental free parameter used in models to "fine-tune" the timing and duration of the reionization epoch that occurred somewhere between 13.4 and 12.7 Gyrs ago (redshifts between 12 > z > 6). However, direct observation of Lyman continuum (LyC) photons emitted below the rest frame \ion{H}{1} ionization edge at 912 �� is increasingly improbable at redshifts z > 3, due to the steady increase of intervening Lyman limit systems towards high z. Thus UV and U-band optical bandpasses provide the only hope for direct, up close and in depth, observations of the types of environment that favor LyC escape. By quantifying the evolution over the past 11 billion years (z < 3) of the relationships between LyC escape and local and global parameters ..., we can provide definitive information on the LyC escape fraction that is so crucial to answering the question of, how did the universe come to be ionized? Here we provide estimates of the ionizing continuum flux emitted by "characteristic" (L_{uv}^*) star-forming galaxies as a function of look back time and escape fraction, finding that at z = 1 (7.6 Gyrs ago) L_{uv}^* galaxies with an escape fraction of 1% have a flux of 10^{-19} ergs cm^{-2} s^{-1} ��^{-1}., Science white paper submitted to NASA Solicitation NNH12ZDA008L: Science Objectives and Requirements for the Next NASA UV/Visible Astrophysics Mission Concepts. 6 pages 4 figures - (abstract abridged)

We present observations of the O VI system at z_abs = 0.495096 toward PKS 0405-123 (z_em = 0.5726) obtained with FUSE and STIS. In addition to strong O VI, with log N(O VI) = 14.47+/-0.02, and moderate H I this absorber shows C III, N IV, O IV, and O V, with upper limits for another seven ions. The large number of available ions allows us to test ionization models usually adopted with far less contraints. We find the ionic column densities cannot be matched by collisional ionization models, in or out of equilibrium. Photoionization models can match the observed column densities, including O VI. If one assumes photoionization by a UV background dominated by QSOs, the metallicity of the gas is [O/H] = -0.15, while a model for the UV background with contributions from ionizing photons escaping from galaxies gives [O/H] = -0.62. Both give [N/O] ~ -0.6 and [C/H] ~ -0.2 to -0.1. The choice of ionizing spectrum is poorly constrained. Multiphase models with a contribution from both photoionized gas (at T~10^4 K) and collisionally ionized gas (at T~(1-3)x10^5 K) can also match the observations giving very similar metallicities. The O VI in this system is not necessarily a reliable tracer of WHIM matter. We do not detect Ne VIII or Mg X absorption. The limit on Ne VIII/O VI < 0.21 (3 sigma), the lowest yet observed. Thus this absorber shows no firm evidence of the "warm-hot intergalactic medium" at T~(0.5-3)x10^6 K thought to contain a significant fraction of the baryons at low redshift. We present limits on the total column of warm-hot gas in this absorber as a function of temperature. This system would be unlikely to provide detectable X-ray absorption in the ions O VII or O VIII even if it resided in front of the brighter X-ray sources in the sky., Accepted for publication in MNRAS. 21 pages. Abstract abridged slightly

Current measurements indicate that the deuterium abundance in diffuse interstellar gas varies spatially by a factor of ~4 among sightlines extending beyond the Local Bubble. One plausible explanation for the scatter is the variable depletion of D onto dust grains. To test this scenario, we have obtained high signal-to-noise, high resolution profiles of the refractory ion TiII along seven Galactic sightlines with D/H ranging from 0.65 to 2.1x10^-5. These measurements, acquired with the recently upgraded Keck/HIRES spectrometer, indicate a correlation between Ti/H and D/H at the >95% c.l. Therefore, our observations support the interpretation that D/H scatter is associated with differential depletion. We note, however, that Ti/H values taken from the literature do not uniformly show the correlation. Finally, we identify significant component-to-component variations in the depletion levels among individual sightlines and discuss complications arising from this behavior., 4 pages; Accepted to Astrophysical Journal Letters

The interstellar thick disks of galaxies contain not only gas, but significant quantities of dust. Most of our knowledge of extraplanar dust in disk galaxies comes from direct broadband optical imaging of these systems, wherein the dust is identified due to the irregular extinction it produces against the thick disk and bulge stars. This observational technique is sensitive to only the most dense material, and we argue much of the material identified in this way traces a cold phase of the interstellar thick disks in galaxies. The presence of a cold, dense phase likely implies the interstellar pressures in the thick disks of spiral galaxies can be quite high. This dense phase of the interstellar medium may also fueling thick disk star formation, and H-alpha observations are now revealing H II regions around newly-formed OB stars associations in several galaxies. We argue that the large quantities of dust and the morphologies of the structures traced by the dust imply that much of the extraplanar material in disk galaxies must have been expelled from the underlying thin disk., 8 pages, 3 figures. To appear in the Proceedings of the Extra-planar Gas Conference, June 7-11, 2004, Dwingeloo (ASP Conf. Series, ed. R. Braun). High resolution PDF version can be found at http://kingpin.ucsd.edu/~howk/Papers/Proceedings/howk2005_dwingeloo.pdf

We present results from an analysis of FUSE and HST/STIS spectroscopy of the z_em=0.57 quasar PKS0405-123. We focus on the intervening metal-line systems identified along the sightline and investigate their ionization mechanism, ionization state, and chemical abundances. We identify six OVI absorbers to a 3sigma equivalent width (EW) limit of 60mA. This implies an incidence dN/dz = 16^{+9}_{-6} consistent with previous OVI studies. In half of the OVI systems we report positive detections of CIII suggesting the gas is predominantly photoionized, has multiple ionization phases, or is in a non-equilibrium state. This contrasts with the general description of the warm-hot intergalactic medium (WHIM) as described by numerical simulations where the gas is predominantly in collisional ionization equilibrium. We have also searched the sightline for the NeVIII doublet (a better probe of WHIM gas with T>10^6K) over the redshift range 0.210^14 cm^-2 is [M/H] ~> -1.5 with large scatter. This significantly exceeds the median metallicity of CIV and OVI systems at z~3 and indicates enrichment of the intergalactic medium over the past 10Gyr. [abriged], Comment: 21 pages, 21 figures (mostly color). Accepted to ApJ, August 12, 2004

We discuss interstellar temperature determinations using the excitation equilibrium of the ^2P levels of Si II and C II. We show how observations of the ^2P_3/2 fine structure levels of Si II and C II (which have significantly different excitation energies, corresponding to ~413 and 92 K, respectively) can be used to limit gas kinetic temperatures. We apply this method to the z=4.224 damped Lyman-alpha system toward the quasar PSS1443+27. The lack of significant absorption out of the SiII ^2P_3/2 level and the presence of very strong C II ^2P_3/2 provides an upper limit to the temperature of the C II*-bearing gas in this system. Assuming a solar Si/C ratio, the observations imply a 2-sigma limit T, Accepted to ApJ Letters; 13 pages, including figures and tables

We present Keck/HIRES abundance measurements and metal-line kinematic profiles of the damped Lyman alpha systems (DLAs) towards the quasars Q0933+733 (z_abs=1.479) and Q0948+433 (z_abs=1.233). These two DLAs have among the five highest HI column densities at any redshift: N(HI)=4.2E21 cm^{-2}. The metal-line data, presented here for the first time, reveal that these DLAs are noteworthy for several other reasons as well. 1) The Q0933+733 DLA exhibits simple kinematic structure with unusually narrow velocity widths as measured from its unsaturated metal lines (delta v=16 km/s). At 2.6% solar, it has the second lowest metallicity at z -1) and a significant alpha-enhancement. The strong metal lines of this DLA have made possible the detection of TiII1910, CoII2012, and MgI2026. 3) We find that the relative gas-phase abundances of both DLAs follow the general trend seen at high redshift, e.g., enhanced Zn/Fe and Si/Fe, and sub-solar Mn/Fe, indicating that there is little evolution in the nucleosynthetic patterns of DLAs down to this epoch. 4) Their high HI column densities imply that these DLAs dominate the column density-weighted cosmic mean metallicity, , of the universe at z, 25 pages. Accepted for publication in the Astronomical Journal

We present FUSE observations of the post-asymptotic giant branch star von Zeipel 1128 (l=42.5, b=+78.7; z=10.0 kpc), located in the globular cluster Messier 3. The FUSE observations cover the wavelength range 905-1187 Ang at ~20 km/s (FWHM) resolution. These data exhibit many photospheric and interstellar absorption lines, including absorption from ions associated with the warm neutral, warm ionized, and highly-ionized phases of the interstellar medium along this sight line. Ionized hydrogen represents >12%, most likely ~45%, of the total hydrogen column along this sight line, most of it associated with the warm ionized phase. The warm ionized and neutral media toward von Zeipel 1128 have very similar gas-phase abundances and kinematics: the neutral and ionized gases in this region of the thick disk are closely related. Strong O VI absorption is seen with the same central velocity as the warm ionized gas, though the O VI velocity dispersion is much higher (sigma=32 km/s). Virtually all of the O VI is found at velocities where lower-ionization gas is seen, suggesting the O VI and WNM/WIM probes are tracing different portions of the same structures (e.g., the O VI may reside in interfaces surrounding the WNM/WIM clouds). We see no evidence for interstellar absorption associated with the globular cluster Messier 3 itself nor with the circumstellar environment of von Zeipel 1128. Neither high velocity cloud absorption (with v(LSR)>125 km/s) nor high velocity-dispersion gas (with sigma~60 km/s) is seen toward von Zeipel 1128. [Abridged], Astrophysical Journal, in press; 39 pages, including 8 figures and 8 tables. Also available at http://kingpin.ucsd.edu/~howk/Papers/

We report Far Ultraviolet Spectroscopic Explorer (FUSE) observations of interstellar O VI absorption in the halo of the Milky Way towards 12 early-type stars in the Large Magellanic Cloud (LMC) and 11 in the Small Magellanic Cloud (SMC). The mean column densities of O VI associated with the Galactic halo towards the LMC and SMC are log = 14.52(+0.10,-0.14) and 14.13(+0.14,-0.20), respectively, where the uncertainties represent the standard deviations of the individual measurements about the means. Significant variations in the O VI column densities are observed over all of the angular scales probed by our observations: 0.5-5.0 degrees towards the LMC and 0.05-3.3 degrees towards the SMC. The maximum factors by which the O VI varies between sight lines towards the LMC and SMC are ~2.8 and ~4.2, respectively. Though low-, intermediate-, and high-velocity clouds are present along most of the sight lines towards the LMC, the variations in O VI column densities are found in all of these O VI absorbing structures. The column density variations trace the structure of the hot ionized medium in the Galactic halo on scales of <, To appear in The Astrophysical Journal. 24 pages including 9 figures and 3 tables

We present the results of a FUSE survey of O VI 1031.93 Angstrom and 1037.62 Angstrom absorption toward 18 OB stars in the Small Magellanic Cloud (SMC). We detect O VI along every sight line. The average value of the O VI column density in the SMC is log =14.53. This value is 1.7 times higher than the average value for the Milky Way halo (perpendicular to the Galactic plane) of log N_perp(O VI)=14.29 found by FUSE, even though the SMC has much lower metallicity than the Galaxy. The column density in the SMC is higher along sight lines that lie close to star-forming regions, in particular NGC 346 in the northern part of the SMC. The standard deviation of the column densities for sight lines outside of NGC 346 is +/- 27%, somewhat lower than the deviation seen in the Milky Way halo. The lowest O VI column densities, log N(O VI)~14.3, occur in the central region and in the southeastern ``Wing'' of the galaxy. Even these low column densities are as high as the Milky Way average, establishing the presence of a substantial, extended component of coronal gas in the SMC. The O VI absorption is always shifted to higher velocities than the main component of lower ionization gas. The O VI line widths are broader than expected for pure thermal broadening. We discuss several mechanisms that may be able to explain the observed properties of the hot gas, including supershells, a galactic fountain, and the infall of gas previously stripped from the SMC by tidal interactions with the Milky Way and the Large Magellanic Cloud. If a fountain produces the hot gas, the mass flux per unit surface area is ~0.02 M_sun/yr/kpc2., Comment: Accepted for publication in the ApJ. 23 pages, 5 figures, 5 tables

We present an atlas of 57 Large Magellanic Cloud (LMC) and 37 Small Magellanic Cloud (SMC) observations obtained with the Far Ultraviolet Spectroscopic Explorer (FUSE) satellite. The atlas highlights twelve interstellar absorption line transitions at a resolution of ~15 km/s. These transitions cover a broad range of temperatures, ionization states, and abundances. The species included are OVI, which probes hot (T~3x10^5 K) ionized gas; CIII and FeIII, which probe warm (T~10^4 K) ionized gas; SiII, PII, CII, FeII, and OI, warm neutral gas; and six different molecular hydrogen transitions, which trace cold (T, 29 pages, 6 figures. Complete Atlas of 94 additional images (~800kB each) is available at http://fuse.pha.jhu.edu/~danforth/atlas Accepted to the ApJS March 2002

I summarize Far Ultraviolet Spectroscopic Explorer (FUSE) observations of interstellar O VI absorption towards 12 early-type stars in the Large Magellanic Cloud (LMC), the closest disk galaxy to the Milky Way. LMC O VI is seen towards all 12 stars with properties (average column densities, kinematics) very similar to those of the Milky Way halo, even though O/H in the LMC is lower by a factor of ~2.5. Sight lines projected onto known LMC superbubbles show little enhancement in O VI column density compared to sight lines towards quiescent regions of the LMC. The O VI absorption is displaced by \~-30 km/sec from the corresponding low-ionization absorption associated with the bulk of the LMC gas. The LMC O VI most likely arises in a vertically-extended distribution, and I discuss the measurements in the context of a halo composed of radiatively-cooling hot gas. In this case, the mass-flow rate from one side of the LMC disk is of the order 1 solar mass/yr., 4 pages, including figures. To appear in the Proceedings of the XVIIth IAP Colloquium: Gaseous Matter in Galactic and Intergalactic Space

We present new, deep optical images (BVI+H-alpha) of the interstellar medium (ISM) far above the plane of NGC 891. These sub-arcsecond images give a direct visual view of two physically distinct ``phases'' of the thick interstellar disk of this galaxy. A dense phase of the thick disk ISM is observed in our BVI images as highly-structured dust-bearing clouds viewed against the stellar light of the galaxy. These structures are traceable to heights |z|=2 kpc from the midplane. Very few highly-structured dust features are present at |z|>2 kpc. The more prominent dust structures have gas masses in excess of 10^5 solar masses, each having visual extinctions well in excess of unity. A warm ionized phase of the high-z ISM is observed through its well-studied H-alpha emission. Our images of the well-studied diffuse ionized medium, to date the highest-resolution observations of this phase of the ISM in NGC 891, show it is relatively smoothly distributed with some filamentary structure superposed on this smooth background. There is little correspondence between the H-alpha emitting material and the absorbing dust structures. These two phases of the multiphase high-z ISM are physically distinct. The H-alpha emission is being heavily extincted in many places by the dense dust-bearing medium. Our H-alpha observations show evidence for several discrete H II regions at large distances from the midplane (to |z|=2 kpc). The presence of these H II regions in the thick disk of NGC 891 suggests that on-going star formation may be present in some of the dense, high-z clouds visible in our images. (Abstract Abridged), Accepted for publication in The Astronomical Journal. 38 pages, including 6 embedded tables and 14 embedded figures. Interested readers are strongly encouraged to download a version with high-resolution figures from http://fuse.pha.jhu.edu/~howk/Papers/papers.html

Recent optical and submillimeter observations have begun to probe the existence of dust grains in the halos of spiral galaxies. I review our own work in this area which employs high-resolution optical images of edge-on spiral galaxies to trace high-z dust in absorption against the background stellar light of the galaxies. We have found that a substantial fraction of such galaxies (>50%) show extensive webs of dust-bearing clouds to heights z>2 kpc. Extraplanar dust in galaxies is statistically correlated with extraplanar diffuse ionized gas, though there is no evidence for a direct, physical relationship between these two phases of the high-z interstellar medium. The dense high-z clouds individually have masses estimated to be >10^5} to 10^6 solar masses. The detailed properties of the observed dust structures suggest the clouds seen in our images may represent the dense phase of a multiphase ISM at high-z. Such dense clouds can have an important effect on the observed light distribution in spiral galaxies. I discuss the effects such high-z dust can have on quantitative measures of the vertical structure of stars and ionized gas in edge-on systems., To appear in "Toward a New Millennium in Galaxy Morphology," eds. D.L. Block, I. Puerari, A. Stockton, & D. Ferreira (Kluwer, Dordrecht). 8 pages; 3 embedded figures. Version with high-resolution figures available at http://fuse.pha.jhu.edu/~howk/Papers/papers.html

We present high-resolution ( 0.5 - 2.0. If Milky Way gas-to-dust relationships are appropriate, then these structures have gaseous column densities N_H > 10(21) cm(-2), with very large masses (>10(5) - 10(6) solar masses) and gravitational potential energies (> 10(51) - 10(52) ergs relative to z=0). The estimated column densities suggest molecular gas may be present, and with the estimated masses allows for the possibility of star formation in these dusty clouds. Recent star formation is the likely cause of the discrete H II regions, in some cases associated with relatively blue continuum sources, observed at heights 0.6 < |z| < 0.8 kpc above the disks of these galaxies. The presence of early-type stars at high-z in these galaxies may be related to the extraplanar dust structures seen in our images., 4 pages; 2 jpeg figures. High resolution figures available at http://www.astro.wisc.edu/~howk/Papers/papers.html#Zermatt To appear in proceedings of the Third Cologne-Zermatt Symposium on The Physics and Chemistry of the Interstellar Medium