The purpose of this dissertation is to expand our understanding of oxygen ion escape to space from Venus and its dependence on extreme solar wind conditions found during interplanetary coronal mass ejections (ICMEs). This work uses in-situ measurements of ions and magnetic fields from the Venus Express (VEX) spacecraft, which has been orbiting Venus since mid-2006. VEX is in a 24 hour elliptical orbit. The ion instrument operates for ~6 hours near the planet, while the magnetometer is always on. In-situ measurements of the solar wind velocity, density, and magnetic field from Pioneer Venus Orbiter (PVO) are also used for comparison with external conditions during the VEX time period. Coronagraph images from solar monitoring spacecraft are used to identify the solar sources of the extreme solar wind measured in-situ at Venus. For interpretation of planetary ions measured by VEX, a magnetohydrodynamic (MHD) model is utilized.

The solar wind dynamic pressure outside of the Venus bow shock did not exceed ~12 nPa, during 2006-2009, while the solar wind dynamic pressure was higher than this for ~10% of the time during the PVO mission. Oxygen ions escape Venus through multiple regions near the planet. One of these regions is the magnetosheath, where high energy pick-up ions are accelerated by the solar wind convection electric field. High energy (>1 keV) O+ pick-up ions within the Venus magnetosheath reached higher energy at lower altitude when the solar wind was disturbed by ICMEs compared to pick-up ions when the external solar wind was not disturbed, between 2006-2007. However, the count rate of O+ was not obviously affected by the ICMEs during this time period. In addition to high energy pick-up ions, VEX also detects low energy (~10-100 eV) O+ within the ionosphere and wake of Venus. These low energy oxygen ions are difficult to interpret, because the spacecraft's relative velocity and potential can significantly affect the measured energy. If VEX ion data is not corrected for the spacecraft's relative velocity and potential, gravitationally bound O+ could be misinterpreted as escaping. These gravitationally bound oxygen ions can extend on the nightside to ~-2 Venus radii and may even return to the planet after reaching high altitudes in the wake. Gravitationally bound ions will lower the total O+ escape estimated from Venus if total escape is calculated including these ions. However, if the return flux is low compared to the total escaping outflow, this effect is not significant.

An ICME with a dynamic pressure of 17.6 nPa impacted Venus on November 11, 2011. During this ICME, the high energy pick-up O+ and the low energy O+ ions were affected. Oxygen ions in the magnetosheath, ionosphere, and tail had higher energies during the ICME, compared to O+ energies when the external solar wind conditions were undisturbed. High energy ions were escaping within the dayside magnetosheath region when the ICME was passing as well as when the solar wind was undisturbed. However, during the ICME passage, these O+ ions had three orders of magnitude higher counts. The low energy O+ during the undisturbed days was gravitationally bound, while during the ICME a portion of the low energy ions were likely escaping. The most significant difference in O+ during the ICME was high energy pickup ions measured in the wake on the outbound portion of the orbit. These ions had an escape flux of 2.5  108 O+cm-2sec-1, which is higher than the average escape flux in all regions of the wake. In addition, the interplanetary magnetic field (IMF) was in a configuration that may have rotated an even higher escape flux O+ away from the VEX orbit. This needs to be confirmed with sampling of other regions in the wake during large ICMEs. A lower bound on the total O+ escape during this event could be ~2.8  1026 to 6.5  1027 O+/sec, which is 2-3 orders of magnitude higher than the average escape flux measured by VEX. Hence, ICMEs could have played a major role in the total escape of O+ from Venus. Considering that the Sun was likely more active (with more ICMEs) early after solar system formation.

The results presented in this dissertation can be used as a guide for future studies of O+ escape at Venus. As we move into solar maximum, Venus will likely be impacted by more large ICMEs. The ICME from the last study of this dissertation was the largest yet measured by VEX, but its 17.6 nPa dynamic pressure is lower than the largest ICMEs during the PVO time period (~ 80 nPa). The work in this dissertation is also relevant to Mars, since Mars interacts with the solar wind in a similar manner and has analogous ion escape mechanisms. The upcoming MAVEN (Mars Atmosphere and Volatile Evolution) mission will launch at the end of 2013 to study the Martian atmosphere, escape processes, and history of volatiles. This mission will have an in-situ ion instrument and magnetometer similar to those used for the studies in this dissertation, so one could conduct similar studies of the oxygen ion escape from Mars during extreme solar wind conditions.

Modern radio astronomy has brought forth an era of data explosion. With advances in instrumentation and computational power, new host of algorithms for data analysis become possible and necessary. Techniques of data reduction and interference rejection are crucial to signal detection for a wide range of scientific questions. In 21\,cm cosmology, a faint and diffuse signal from primordial hydrogen requires extracting every last bit of sensitivity from the instruments. In search of fast radio burst and signals of extra-terrestrial technology we need to sort through large volumes of data while effectively distinguishing candidate signals from noise and interference. Overcoming these challenges requires novel intuition of the experiments as well as applications of state-of-the art methods of statistics.

Redundant arrays are a kind of interferometers designed to detect the 21cm power spectrum. The spectrum itself is naturally linked to the square of the visibilities, giving rise to a set of theoretical formulations that skips the need for forming images. Such formulation, however, does not naturally include all sensitivity of the interferometer. Thus in this thesis we develop a technique to complete this formulation and provide the remaining sensitivities.

Fast radio bursts (FRB) are a type of millisecond-duration transient signal seen in spectrograms with a quadratically dispersed shape. At the time of the writing of this thesis, the physical nature of the sources is yet unknown. One of the repeating sources is FRB121102. Using deep learning, we detect the most number of signals from FRB121102 during a single observation. The 93 pulses provide new wealth of statistical properties, most notably periodicity. Whether the emission is periodic provides an insightful look on the nature of the source. With our novel method of periodicity search, we show that the received pulses are likely not periodic with any period longer than 10 milliseconds.

Radio frequency search for extraterrestrial intelligence (SETI) is one of the most open-ended questions in astronomy. With properties of the candidate signal unknown, SETI needs intelligent methods that automatically extracts information from large volumes of data. In parallel with new developments in radio astronomy, machine learning techniques, in particular deep learning has successfully revolutionized applications in computer vision, speech processing, text processing, and signal processing. In this thesis, we demonstrate a range of data processing techniques, with focus on deep learning, to signal detection in SETI. We showcase a blind detection for repeating signals, a semi-supervised representation learning framework, as well as a generative model for anomaly detection and interference rejection.

Giant planets are characterized by their predominant gaseous composition, with hydrogen and helium as the major constituents. The presence of trace gases in the atmosphere has two effects. Condensible trace gases form clouds, restricting visible and near-infrared observations mostly to the top-most cloud layer. This severely limits our understanding of the atmospheric composition and global circulation, both of which are hidden below the clouds. However, their interaction with radiation also provides a window through the clouds. We can use spectroscopy to retrieve the distribution below the clouds and to investigate the three-dimensional structure of the atmosphere through mid-infrared and radio observations.

One of the key trace gases in the atmospheres of giant planets is ammonia (NH3). Ammonia constitutes a small fraction of the atmosphere (approximately 300 parts per million), but it serves as a valuable tracer for the background circulation by examining its distribution and temporal evolution. In the radio regime, the spectral absorption features of ammonia are highly pressure-broadened, allowing us to probe depths ranging from tens to thousands of bars. Through comparisons of radio observations with model atmospheres, where we can manipulate the distribution of trace gases and temperature profiles, we can constrain the atmospheric properties of Jupiter to hundreds of bars. These comparisons provide valuable insights into the processes shaping the atmosphere.

Prior to the arrival of the Juno mission, radio observations had already indicated that Jupiter’s atmosphere is neither latitudinally nor vertically well-mixed below the cloud deck. Based on these earlier observations, it was believed that the water condensation layer around 6 bars acted as a barrier for atmospheric circulation, below which the atmosphere was considered fully convective and well-mixed. However, the presence of synchrotron radiation, the dominant source of radiation at lower frequencies, limited our understanding to altitudes above the water condensation layer.

The arrival of the Juno mission in the Jovian system has enabled us to overcome this lim2 itation by descending below the synchrotron radiation belts (peaking at 1.5 jovian radii). Throughout my Ph.D. research, I have developed tools to independently calibrate and process instrument data from radio observatories, focusing primarily on NASA’s Juno Microwave Radiometer (MWR) and NRAO’s Very Large Array (VLA). The open-source pipelines I developed retrieve relevant observational quantities (brightness temperature and its dependence on the emission angle) along with their uncertainties. Subsequently, I developed and compared models with observations, which allowed me to study the distribution of ammonia on Jupiter. The enhancement in nitrogen (2.66+0.27 −0.17 solar) in Jupiter’s deep atmosphere was found to be lower than previously estimated, and consistent with the noble gases in the atmosphere as obtained by the Galileo probe.

A more surprising discovery was the depth to which the atmosphere appears to be depleted (20-30 bar), much deeper than can be explained by dynamics or conventional microphysics. This finding highlights that Jupiter might have a stably-stratified troposphere. The distribution of ammonia revealed that zones and and belts on Jupiter can be traced below the cloud deck; in the tropical regions extending down to 20 bars, while the subtropics and mid-latitudes are confined to the upper few bars. Deeper in the atmosphere, around 30 bars, the signal reverses, showing enhanced belts compared to the depleted zones. This distribution could be explained by a stacked circulation cell model, where the middle and lower troposphere cells have opposite circulation directions.

By examining the longitudinal structure of the atmosphere using data from both VLA and MWR, I was able to analyze the weather patterns on Jupiter. The majority of the variability in Jupiter’s atmosphere is confined to the upper few bars, affirming earlier theories and observations that weather is a shallow phenomenon on the gas giant. Even in the most dynamically active region (the North Equatorial Belt), the weather extends only to approximately the water condensation layer. The only three features I identified capable of affecting the lower troposphere were an anti-cyclone, an ammonia plume as part of the trapped equatorial Rossby wave, and the signature of precipitation, which appeared to deposit ammonia below the water condensation layer.

The breakthrough in understanding came from studying the impact of a giant outbreak in the South Equatorial Belt. By comparing the MWR observations of the outbreak with orbit-averaged observations, I demonstrated that the storm affected the deep troposphere down to approximately 20 bars. These are the first observations showing that storms deplete the upper atmosphere in ammonia and deposit ammonia, while also cooling the lower troposphere at much greater depths than previously anticipated on Jupiter. The mechanism through which the atmosphere is depleted favors supercooled water, rapidly updrafted in storms, absorbing ammonia to form water-ammonia hail, which is subsequently deposited around 15-25 bars. This process shows how small- and large-scale processes interact to shape the atmosphere of Jupiter. This work has shown that the gas giant atmosphere are not as convective as previously thought, and are controlled by the interaction of both local microphysics and the background large-scale circulation.

The atmospheric chemistry and dynamics of the ice giants are critical to comprehending planetary diversity within our Solar System and beyond. The bulk composition of these planets, which can be investigated through the composition of their atmospheres, is intimately tied to their formation in the circumstellar disk. In this dissertation, I take a multi-wavelength approach to studying chemistry and large-scale circulation in the atmosphere of Neptune, one of the two local examples of ice giant planets.

In equilibrium, the molecule carbon monoxide (CO) should be confined to the warm interiors of the giant planets. Its presence in the upper atmosphere, therefore, indicates disequilibrium processes at work: either rapid transport from deeper levels where the molecules are thermodynamically stable, or production high in the atmosphere as a result of infall of material from the planet's environment. Using millimeter-wave interferometry with the Combined Array for Research in Millimeter-wave Astronomy (CARMA), I observe Neptune in two rotational transitions of CO. Radiative transfer modeling yields a CO profile with 0.1^+0.2_-0.1 parts per million (ppm) of CO in the troposphere, and 1.1^+0.2_-0.3 ppm in the stratosphere. The stratospheric abundance implies an infall rate of oxygen-bearing material of 0.5–20×10⁸ cm^-2 s^-1 to Neptune's upper atmosphere, which is consistent with supply by (sub)kilometer-sized comets. I also revisit the calculation of Neptune's internal oxygen abundance using revised calculations for the CO→CH₄ conversion timescale in the deep atmosphere (Vissher & Moses 2011), in the context of my derived CO profile. The best-fit solution of 0.1 ppm of CO in the troposphere implies a global O/H enrichment of at least 400, and likely more than 650 times the protosolar value. This is one order of magnitude greater than Neptune's observed carbon enrichment relative to solar. However, the CO profile is also consistent with 0.0 ppm of CO in the troposphere, in which case no over-enrichment of oxygen in the interior is required.

Maps of Neptune in and near the CO (2–1) line from CARMA show spatial variations in the intensity at the 2–3% level. Variations at frequencies in the CO line are consistent with variations in zonal-mean temperature near the tropopause. At continuum wavelengths, I observe a gradient in the brightness temperature, increasing by 2–3 K from 40°N to the south pole. This corresponds to an opacity decrease of about 0.3 (30%) near the south pole at altitudes below 1 bar, or a factor of 100 decrease in the opacity at altitudes below 4 bar. A global circulation pattern in which moist air rises at mid- southern and northern latitudes and dry air subsides near the equator and south pole is consistent with the south polar brightening observed in the millimeter.

At near-infrared wavelengths, observations are sensitive to sunlight reflected from clouds and hazes in the upper atmosphere. These clouds act as tracers of atmospheric dynamics, which aid in the determination of Neptune's large-scale circulation. Near-infrared images of Neptune from the W.M. Keck II telescope from July 2007 show that the unresolved cloud feature typically observed within a few degrees of Neptune's south pole had split into a pair of bright spots. A careful determination of disk center places the cloud centers within 2 degrees of, but not directly at, Neptune's south pole. When modeled as optically thick, perfectly reflecting layers, the two features are found to be in the troposphere, at pressures greater than 0.4 bar. Images with comparable resolution taken two days later reveal only a single feature near the south pole. The changing morphology of these circumpolar clouds suggests they may form in a region of strong convection surrounding a Neptunian south polar vortex.

Additional near-infrared observations were performed with the OSIRIS integral-field spectrograph on the Keck telescope. I present three-dimensional data cubes covering more than 90% of the visible hemisphere of Neptune, with a spatial resolution of 0.035'' per pixel and spectral resolution of R∼3800, in the H (1.47–1.80 μm) and K (1.97–2.38 μm) broad bands. In my preliminary radiative transfer analysis of these data, I find that the observed spectra are generally well fit by models with three cloud layers: a stratospheric haze layer, a tropospheric haze layer and a low albedo, optically thick bottom cloud at 2 bar. Models are consistent with a north-south hemispheric asymmetry in the properties of the clouds, with both bright and dark features in the north at higher altitudes than equivalent regions in the south. The range of derived altitudes between the highest and deepest features is nearly 5 atmospheric scale heights. The highest concentration of features can be found within a bright band extending from 30 to 45°S. I find that these features can all be fit by hazes at the same altitude, by varying only the haze particle densities. In contrast, I find that two features near 60°S appear to be at very different altitudes, even though they are at the same latitude.

There is a growing need for methodologies that integrate machine learning with scientific domain expertise, enabling the construction of interpretable models while contextualizing rare phenomena within the broader scientific landscape. With the exponentially growing influx of publicly available Earth observation data, such methods can be deployed to help answer open questions in Earth's cryosphere, which will enable better constraints of future sea level rise in the coming century.

This research develops scientifically contextualized computer vision methods in low-shot learning regimes, helping to address critical questions at the ice-bed and ice-ocean interfaces. I focus on the classification of rare phenomena that are characterized by prehistoric and contemporary ice sheets. I develop a scientifically-driven filtering method to automate the detection of subglacial bedforms formed during the last glacial maximum, which were shaped by the dynamics of the ice sheets that once flowed above them. These bedforms compose ~2% of the overall training set, and the proposed prefiltering approach can be modularly inserted into exisiting pipelines, enabling the automatic detection of these bedforms from publically available digital elevation models with up to 94% accuracy.

I present a concise tiling strategy that I developed to prepare large satellite imagery for use with machine learning algorithm training on GPUs. Tiling is the industry standard, but previous methods that attempted to preserve semantic context created redundancies within the training data, altering the model outcomes. By uniquely permuting every extension of the dataset, I eliminate redundancies, and find that, with distinct transformations, I can extend the dataset further than previous approaches, and that this extension doesn't alter the structure of the training data. Applying this preprocessing step alone, without any changes to model architecture or optimization, improves the performance on underrepresented classes by up to 15.8%.

I use this method to prepare manually labeled imagery to search for persistent polynyas, which are a rare phenomena along the western coast of Antarctica. Polynyas are areas of open water within the sea ice, and when opened thermodynamically by warm water plumes that arise from under the marginal ice shelves, they can persist in the same location from year to year. They also offer a surface view into subsurface ice-ocean interactions that would be otherwise hard to monitor in satellite imagery. However their small size and relative rarity, means that these polynyas make up a fractional portion of the overall scene in satellite imagery and are easily confounded with other areas of open water which occur more frequently when off-the-shelf machine learning methods are applied. Due to their physics of their formation pathways, such polynyas typically occur right at the ice front interface, and I use this geometric constraint to build a geophysically contextualized objective into the loss function of existing object detection architectures, allowing for the rapid detection of the persistent polynya census in the Amundsen Sea embayment. I recover all eleven previously characterized polynyas in the region, and find eight new polynyas. While this approach was specified in our particular model to the geomorphometry that informs persistent polynya formation, such an approach is easly generalized to aid in the detection of any underrepresented target for which prior knowledge of semantic contextualizations governs the geometry of the scene.

Uranus and Neptune are representatives of the `ice giants', one of the most common classes of exoplanets \citep{Fressin2013}. Thousands of exoplanets have been discovered thanks to the \textit{Kepler} mission, and soon the James Webb Space Telescope will characterize their atmospheres in unprecedented detail. Such work will rely on the observations, techniques, and analysis used to study the Solar System's gas giants. However, in many ways our own ice giants remain poorly understood. In this dissertation, I use multi-wavelength observations of Neptune to better constrain the bulk properties and dynamic patterns within the planet's upper atmosphere.

At visible and near-infrared wavelengths, sunlight is reflected off the cloud tops and hazes populating the upper atmospheres of the giant planets. Bright cloud features can be tracked to extract velocities. By doing this over many latitudes, a global velocity field called the `zonal wind profile' can be made. Here, I present zonal wind profiles for Jupiter and Neptune. These are constructed from \textit{Hubble Space Telescope} WFC3 global maps of Jupiter taken between $2009-2016$ and Keck NIRC2 images of Neptune taken in the H-band ($1.4-1.8 \mu$m) and Kp-band ($2.0-2.4 \mu$m) in 2013 and 2014.

I show that Jupiter's zonal wind profile is stable throughout the observed period, apart from variations on the order of 10 m/s at the $24^{\circ}$N Northern Temperature Belt (NTB). These variations arise during periodic plume outbreaks at the NTB and are coupled to a decrease in the albedo. These findings suggest that material, normally unseen, is dredged upward due to these plumes. If plumes are a signature of deeper activity, the decrease in velocity we see at the NTB during outbreaks may be evidence of vertical wind shear.

I also find evidence of vertical wind shear at Neptune's equator, with the H-band zonal wind profile offset eastward by 100 m/s at the equator relative to the Kp-band profile. I apply a new thermal wind equation applicable at the equator to reconcile this observed vertical wind shear with Neptune's horizontal thermal and composition profiles. In order to match \textit{Voyager}/IRIS derived temperatures \citep{Fletcher2014}, the equator must be enriched in methane compared to the mid-latitudes at pressures greater than 1 bar. I discuss the implications of this finding with regards to global dynamics and compare and contrast to the other giant planets.

Radio wavelengths probe below the visible cloud deck. I analyze maps of Neptune taken with the Atacama Large Millimeter/Submillimeter Array (ALMA) and extended Very Large Array (VLA) to constrain Neptune's deep opacity sources. The opacity source at radio wavelengths is dominated by H$_2$S and NH$_3$ as well as the the collision-induced absorption of H$_2$ with H$_2$, He, and CH$_4$. Clear brightness temperature variations are present across Neptune's disk caused by variations in these trace gases. These observations are the first to achieve the sensitivity, resolution, and wavelength coverage required to simultaneously extract the abundance profiles of H$_2$S, CH$_4$, and NH$_3$. I retrieve disk-average properties assuming both wet and dry adiabats. The disk-averaged data are consistent with profiles where trace gases are enriched by $30\times$ their protosolar value, apart from NH$_3$ which is $1\times$ its protosolar value.

In both the ALMA and VLA maps, I identify seven distinct latitudinal bands with discrete transitions in the brightness temperature. I use the radiative transfer code Radio-BEAR to generate model spectra of Neptune's brightness temperature as a function of temperature and composition. I find best-fitting parameters to the H$_2$S, NH$_3$, and CH$_4$ abundance profiles in each of the seven identified latitude bands using $\chi^2$-statistics and MCMC retrievals. Of note, the equator is more complicated than expected. Trace gases are enriched in the $2-12^{\circ}$N band compared to neighboring latitudes. Here, the best-fit deep CH$_4$ abundance is $45\times$ the protosolar value (or 2.2$\%$ mixing ratio). H$_2$S is $30\times$ solar (or $7\times10^{-4}$ mixing ratio) and supersaturated at the H$_2$S-ice cloud formation. I relate these findings to my near-infrared work and present a new schematic of Neptune's global circulation structure.

A central goal of planetary science is the creation of a framework within which the properties of each solar system body can be understood as the product of initial conditions acted on by fundamental physical processes. The solar system’s extreme worlds - those objects that lie at the far ends of the spectrum in terms of planetary environment - bring to light our misconceptions and present us with opportunities to expand and generalize this framework. Unraveling the processes at work in diverse planetary environments contextualizes our understanding of Earth, and provides a basis for interpreting specific signatures from planets beyond our own solar system. Uranus and Io, with their unusual planetary environments, present two examples of such worlds in the outer solar system.

Uranus, one of the outer solar system’s ice giants, produces an anomalously low heat flow and orbits the sun on its side. Its relative lack of bright storm features and its bizarre multi-decadal seasons provide insight into the relative effects of internal heat flow and time- varying solar insolation on atmospheric dynamics, while its narrow rings composed of dark, macroscopic particles encode the history of bombardment and satellite disruption within the system.

Jupiter’s moon Io hosts the most extreme volcanic activity anywhere in the solar system. Its tidally-powered geological activity provides a window into this satellite’s interior, permitting rare and valuable investigations into the exchange of heat and materials between interiors and surfaces. In particular, Io provides a laboratory for studying the process of tidal heating, which shapes planets and satellites in our solar system and beyond. A comparison between Earth and Io contextualizes the volcanism at work on our home planet, revealing the effects of planetary size, atmospheric density, and plate tectonics on the style and mechanisms of geological activity.

This dissertation investigates the processes at work on these solar system outliers through studies of Uranus’ atmosphere and rings and of Io’s thermal activity. I show that Uranus’ rings are spectrally flat in the near-infrared, setting them apart from all other ring systems in the solar system. I investigate the vertical profile of species in Uranus’ atmosphere, and demonstrate evidence for seasonal trends in the upper atmosphere on decadal timescales.

Based on a large high-cadence dataset of Io’s volcanism obtained with adaptive optics over 100 nights, I show that the thermal timelines of Io’s volcanoes indicate at least two distinct classes of eruption. The asymmetric spatial distribution of Io’s volcanic heat flow suggests additional mechanisms at work modulating the effects of tidal heating. I present the detection of one of the most powerful eruptions ever seen on Io, which I use to derive a eruption temperature of >1300 K, consistent with a highly mafic magma composition. Geophysical modeling of the thermal timeline of Loki Patera, a distinctive volcanic feature on Io, indicates low lava thermal conductivities also consistent with a highly-mafic silicate composition. Ultra-high-resolution thermal mapping of this patera reveals a multi-phase volcanic resurfacing process that hints at the plumbing system underlying this massive volcanic feature.

The results presented here are founded on near-infrared observations of unprecedented resolution in the spatial, spectral, and temporal domains. The interpretation of the data utilizes rigorous statistical techniques to draw meaningful conclusions. In addition to the scientific impact of the findings, this work therefore also pioneers specific ground-based tele- scope capabilities and analysis tools, and demonstrates their utility to solar system science. Chapter 2 presents the first high-resolution spectra of Uranus’ rings. Chapter 3 introduces Markov Chain Monte Carlo simulations into ice giant atmospheric radiative transfer model- ing, permitting a rigorous analysis of parameter uncertainties and correlations. Chapters 4-7 present results from the first multi-year, high-cadence ground-based observing campaign to study Io’s volcanism with sufficient spatial resolution to directly resolve individual volcanoes. The thermal timelines of these volcanoes provide unprecedented insight into the variability and distribution of Io’s volcanism over a wide range of timescales. Chapter 7 uses geometric arguments to deduce topography of a volcanic feature on Io based on observations at a range of viewing angles. Finally, Chapter 8 presents the first ground-based observations to map a thermal feature on Io at a spatial resolution of ∼10 km on Io’s surface, derived from the first mutual satellite occultation event to be observed with adaptive optics on a dual-telescope interferometric system. These techniques can all be expanded and applied to these and other targets in future near-infrared studies.

This dissertation is devoted to expanding our understanding of the solar wind structure in the inner heliosphere and variations therein with solar activity. Using spacecraft observations and numerical models, the origins of the large-scale structures and long-term trends of the solar wind are explored in order to gain insights on how our Sun determines the space environments of the terrestrial planets.

I use long term measurements of the solar wind density, velocity, interplanetary magnetic field, and particles, together with models based on solar magnetic field data, to generate time series of these properties that span one solar rotation (~27 days). From these time series, I assemble and obtain the synoptic overviews of the solar wind properties.

The resulting synoptic overviews show that the solar wind around Mercury, Venus, Earth, and Mars is a complex co-rotating structure with recurring features and occasional transients. During quiet solar conditions, the heliospheric current sheet, which separates the positive interplanetary magnetic field from the negative, usually has a remarkably steady two- or four-sector structure that persists for many solar rotations. Within the sector boundaries are the slow and fast speed solar wind streams that originate from the open coronal magnetic field sources that map to the ecliptic. At the sector boundaries, compressed high-density and the related high-dynamic pressure ridges form where streams from different coronal source regions interact. High fluxes of energetic particles also occur at the boundaries, and are seen most prominently during the quiet solar period. The existence of these recurring features depends on how long-lived are their source regions.

In the last decade, 3D numerical solar wind models have become more widely available. They provide important scientific tools for obtaining a more global view of the inner heliosphere and of the relationships between conditions at Mercury, Venus, Earth, and Mars. When I compare the model results with observations for periods outside of solar wind disturbances, I find that the models do a good job of simulating at least the steady, large-scale, ambient solar wind structure. However, it remains a challenge to accurately model the solar wind during active solar conditions. During these times, solar transients such as coronal mass ejections travel through interplanetary space and disturb the ambient solar wind, producing a far less predictable and modelable space environment. However, such conditions may have the greatest impact on the planets - especially on their atmospheres and magnetospheres. I therefore also consider the next steps in modeling, toward including active conditions.

We consider four modeling efforts to determine the interior structure of the Moon. This includes a geophysical forward model capturing magnetic induction from conducting layers within a vacuum, a plasma induction model capturing the kinetic plasma environment within the wake cavity around a conducting Moon, a multidisciplinary modeling analysis including geochemical, geophysical, and geodynamical considerations, and forward calculate the seismic free oscillations of the Moon.

Electromagnetic Sounding constrains conducting layers of the lunar interior by observing variations in the Interplanetary Magnetic Field disturbing the Moon. We employ the time domain transfer function method locating transient events observed by two magnetometers near the Moon. We consider ARTEMIS and Apollo magnetometer data. We demonstrate our forward model passes benchmarking analyses and solves the magnetic induction response for any input signal as well as any conductivity profile. Plasma hybrid models use the upstream plasma conditions and IMF to capture the wake current systems formed around the Moon. The plasma kinetic equations are solved for ion particles with electrons as a charge-neutralizing fluid. These models accurately capture the large-scale lunar wake dynamics for a variety of solar wind conditions (ion density, temperature, solar wind velocity, and IMF orientation). This method and modeling results are applicable to any airless body with a conducting interior, interacting directly with the solar wind in the absence of a parent body magnetic field as well as any two-point magnetometer constellation.

For our third analysis, we take a multidisciplinary approach to constrain key parameters of the lunar structure. We use an ensemble of 1D lunar compositional models with chemically and mineralogically distinct layers, and forward calculate physical parameters. We employ a grid search and a differential evolution genetic search algorithm to extensively explore model space, where the thickness of individual compositional layers varies. We find that the proposed partially molten layer within the deep mantle may have been formed by the overturn of an ilmenite rich layer, formed after the crystallization of a magma ocean. Moreover, only a narrow range of core radii are consistent with the mass and moment of inertia constraints. For the models that fit the observed mass and moment of inertia, we compare the deviation in the chemistry to that published, and calculate 1D seismic velocity profiles, the majority of which compare well with those determined by inverting the Apollo seismic data.

The technology used to observe solar system planets continues to advance, increasing the spatial resolution, time cadence, and spectral coverage of observations and permitting the discovery of new phenomena in the atmospheres of worlds near and far. Comparative planetology is a powerful framework to help characterize and explain these newly-discovered phenomena: the similarities and differences between the composition, structure, and dynamics of planetary atmospheres test the robustness of the theories we use to understand and predict the weather and climate of atmospheres. This dissertation uses a comparative planetology framework to understand the physical processes underlying new observations of the atmospheres of Uranus and Neptune and new statistical analyses of observed storms on Earth. Although qualitatively dissimilar, the atmospheres of Uranus, Neptune, and Earth lie in a similar fluid-dynamical regime in many respects, allowing much of our physical intuition to translate across those three atmospheres. In particular, this dissertation focuses on using remote-sensing data to probe processes relevant to cloud formation and precipitation.

Higher bulk densities and strong enrichment in carbon and oxygen distinguish Uranus and Neptune from the gas giants Jupiter and Saturn, placing them into their own category, the ice giants. Exoplanet statistics reinforce this distinction: a gap in the size distribution of known exoplanets has been observed between the Jupiter-sized and Neptune-sized exoplanets. Uranus and Neptune are therefore of primary importance for understanding the different types of worlds that fill our galaxy; however, their distance from Earth also makes them very challenging to study in detail. The first three chapters of this dissertation tackle that observational challenge.Using near-infrared data from Keck Observatory and optical data from the Hubble Space Telescope (HST), I characterize a newly-observed storm at the equator of Neptune, including derivations of drift rates from cloud tracking and cloud-top pressures from radiative transfer modeling. I then apply the physics of deep convection and planetary waves to infer the drivers of the storm. Making use of millimeter-wavelength observations from the Atacama Large (sub)-Millimeter Array (ALMA) and mid-infrared observations from the Very Large Telescope (VLT), I observe the thermal component of the Uranian ring system for the first time. I apply the radiative transfer equation to the ring particles to derive their surface temperature and show that they rotate slowly compared to their radiative cooling time. I use the same dataset in conjunction with centimeter-wavelength Very Large Array (VLA) data to study the deep troposphere of Uranus from 1-50 bar. Applying intuition based on Earth's Hadley-Walker circulation, I link the observed enrichment and depletion of volatile species in Uranus's deep troposphere to Uranus's planetary-scale circulation.

Although the instrumentation and radiative transfer models are similar between observations of Earth and the ice giants, the technical challenges that arise in observing these bodies are somewhat opposite. Earth observations exemplify a data-rich regime in which millions of new data points are recorded every hour at spatial resolutions of order 1 kilometer, whereas ice giant studies are comparatively data-starved: $\sim$10 images at 100-kilometer resolution may constitute a full dataset. The second half of this dissertation tackles the data-rich observational challenge, concerning itself with the observed statistics of extratropical extreme storms on Earth from remote-sensing data. Using Doppler radar data, I show that the return values of extreme precipitation are strongly autocorrelated over the Eastern and Midwestern United States up to scales of $\gtrsim$100 kilometers, and that rain gauges can accurately estimate extreme precipitation only if interpolated with a particular eye toward capturing extremes. I then compile a suite of hourly data at 8-km resolution over the Eastern and Midwestern United States that simultaneously probes the synoptic-scale dynamical context of storms, the structure of clouds within storms, and the precipitation produced by storms. I demonstrate the utility of this unique data combination by examining the precipitation efficiency of storms, showing that precipitation efficiency increases with equivalent potential temperature over the contiguous United States (CONUS) in both the warm and cool seasons.

Taken as a whole, this dissertation showcases the ways in which recent advances in remote sensing technology and computational power can be applied to study atmospheres across the solar system, with a particular focus on cloud and precipitation processes.The types of questions that can be answered in the data-rich regime underscore the massive potential for advances in our understanding of the Solar System planets (and exoplanets) as data from new observatories and missions arrive in the next decades.