UC Santa Cruz

Since the beginning of the modern telescope, astronomers have thought of new surveys and methods to study astrophysical phenomena. In this dissertation, I present the Swope Supernova Survey, a low-redshift photometric survey at Las Campanas Observatory, Chile, detailing its motivation, methodology, and significant contributions to transient astrophysics. I also highlight my vital contributions to the survey and science enabled. Since its inception in 2016, the survey has established itself as a critical resource for the study of transients below +30◦ declination, covering a wide wavelength range (u to i band), precise calibration, and high observing cadences. I specifically focus on the first Type Ia Supernova (SN Ia) data release, an effort that I led to provide over 100 high-cadence light curves in five photometric bands. This dataset enhances low-redshift SN Ia samples and opens the path for future work that will significantly contribute to SN cosmology. Finally, I introduce a novel parametrization of i-band light-curve diversity. I present the ∆m1 − ∆m2 parameter, which captures differences between the data and model at the i-band secondary maximum and minimum. Strong correlations are identified between this parameter and key spectral features, such as Ca II pEW0 and Si II v0, highlighting the role of spectral variations in shaping i-band light curves. This work also shows how these variations impact SN Ia composite spectra and synthetic photometry, revealing limitations in the widely used SALT3 SN Ia model. This dissertation highlights the importance of combining photometric and spectroscopic analyses to advance our understanding of SNe Ia, further exploring connections between SN Ia spectral features, i-band light-curve morphology and diversity, physical processes, environmental dependencies, and the accuracy of SNe Ia as precise cosmological distance indicators.

A search for the Standard Model Higgs boson produced in association with a high energy photon is performed using 133 fb−1 of pp collision data collected at √s = 13 TeV with the ATLAS detector at the Large Hadron Collider at CERN. The H + photon final state is particularly promising to study because the photon requirement reduces the multijet background, and the bb final state is the dominant decay mode of the Higgs boson. Event selection requirements isolate vector boson fusion Higgs production, the dominant production mode in this channel. Several improvements enhance the search sensitivity compared to previous measurements, including better background modeling and characterization, use of a dense neural network classifier, and an updated signal extraction strategy adopting a binned-likelihood fit directly to the classifier discriminant. These advancements result in a Higgs boson signal strength measured as 0.2 ± 0.7 relative to the Standard Model prediction. This corresponds to an observed significance of 0.3 standard deviations, compared to 1.5 standard deviations expected signal significance.

Like other marine mammals, True seals (Family Phocidae) rely on acoustic cues for orientation, communication, and prey and predator detection. Because of their amphibious life histories, the auditory systems of seals must operate efficiently both in air and water—environments with very different physical characteristics. While all seals exhibit common evolutionary traits related to hearing, the extent of auditory adaptations varies between phylogenetic lineages and, in some cases, may differ among species. The functional significance of these differences remains to be resolved. The most complete dataset describing amphibious hearing in seals is for the Phocinae subfamily (most temperate and polar phocid species of the Northern Hemisphere). There are few hearing data available for seals from the Monachinae subfamily (the Southern Ocean seals, monk seals, and elephant seals). However, the limited evidence suggests potential subfamily-level differences in hearing. Additional audiometric measurements are needed within the Monachinae lineage of seals to inform our understanding of auditory adaptations from an evolutionary perspective.

The first two chapters of this dissertation aim to expand knowledge of amphibious hearing in seals—particularly from the lesser known Monachinae lineage—by utilizing classic behavioral methods with two individual Hawaiian monk seals (Neomonachus schauinslandi) conditioned to voluntarily participate in hearing trials. These efforts generated and validated the first terrestrial audiogram, provided the first auditory masking measurements, and resolved discrepancies between two prior underwater hearing profiles for monk seals. The findings suggest reduced terrestrial hearing sensitivity may be related to physiological differences in soft tissue within the peripheral auditory system among seal species, which could inhibit the reception of airborne sound. Together, the results confirm that the hearing abilities of monk seals differ from those of related species and are informative for evolutionary considerations of hearing in seals.

From an applied perspective, these hearing data suggest that terrestrial communication is limited for the species. However, a lack of data describing the amplitude of Hawaiian monk seal airborne vocalizations has precluded any communication range estimates. For Chapter 3, I describe the spectral characteristics of and provide the first source level measurements for low-frequency calls emitted by this species in air. These amplitude and spectral data are combined with hearing thresholds and representative ambient noise levels to estimate the distances over which these seals can effectively communicate with conspecifics. Findings suggest that terrestrial communication is limited by the poor hearing sensitivity and moderate vocal amplitudes of the species and is further constrained by ambient noise in the environment.

This series of audiometric measurements advances knowledge of acoustic sensitivity in an endangered species, contributes comparative information about hearing for a data-poor marine mammal lineage, and increases our understanding of the evolution of hearing in the amphibious true seals. Finally, by combining hearing data with information about sound production, we can better understand the acoustic communication system of Hawaiian monk seals, ultimately supporting conservation and management efforts for this endangered species.

This thesis presents a unified classification of semisimple Lie algebras and Kac-Moodyalgebras through their shared foundation in Cartan matrices and Dynkin diagrams. Motivated by the systematic classification of finite-dimensional Lie theory and inspired by Professor Chongying Dong’s insight, this work systematically explores the algebraic and geometric frameworks underpinning both classifications. For semisimple Lie algebras, we establish the classification via root systems and Dynkin diagrams, emphasizing Cartan’s criterion and the Killing form. Extending these principles, Kac-Moody algebras are constructed through generalized Cartan matrices, revealing infinite-dimensional symmetries critical to modern theoretical physics. By bridging finite and infinite dimensions, this thesis highlights applications in string theory, conformal field theory, and quantum gravity, while demonstrating how combinatorial tools like Dynkin diagrams unify disparate algebraic structures. The representation theories of both algebras are examined, culminating in the Weyl and Weyl-Kac character formulas, which underpin physical systems from atomic spectra to vertex operator algebras.

Significant work has been done in the field of extreme analysis in the form of generalization of the univariate generalized Pareto distribution to a multivariate setting. We consider the constructive definition of the multivariate Pareto that factorizes a Pareto random vector into independent radial and angular components; the former following a Pareto distribution, the latter following a distribution with no closed form with support on the surface of the positive orthant of the L-infinity-norm unit hypercube. In this document, we propose a method of inferring this angular distribution, as a realization of a Bayesian non-parametric mixture of independent random gamma vectors, projected onto an arbitrary L-p-norm unit hypersphere; the support of which will approach the support of the angular component as p goes to infinity. We explore applications of this BNP mixture of projected gammas in characterizing the dependence structure of extremes; the motivating example of such we examine is the integrated vapor transport, data pertaining to an atmospheric river transporting moisture from the Pacific ocean across California. We observe clear but heterogeneous geographic dependence. Second, we consider the application of the BNP mixture of projected gammas to a novelty detection setting, developing novelty scores appropriate to the support. To expand the applicability of our methods, we develop a categorical data model, and consider the extension of the angular novelty scores to categorical, and mixed data settings. We find that our model and scores compare favorably to canonical novelty scores on canonical novelty detection datasets. Finally, we seek to understand the limitations of BNP mixture of projected gammas, by attempting to apply the model at a large scale---applied to storm surge data at specified locations, as simulated under the Sea, Lakes, and Overland Surges due to Hurricanes (SLOSH) model. We observe issues in model fidelity, in terms of recovering the marginal distributions, or capturing the dependence structure in a highly multivariate setting. We observe that as dimensionality increases, the number of extant clusters decreases. To ameliorate this loss of granularity, a regression model is proposed, that invokes a low-dimensional representation of the output space. We use these models to explore storm surge at sites of critical infrastructure in the Delaware Bay watershed.

RNA is central to life, acting as a bridge between genetic information and functional proteins. This dissertation explores two key aspects of RNA biology: the technical biases introduced by poly(A) selection in RNA sequencing and the mechanistic processes of Nonsense-Mediated mRNA Decay (NMD) in Caenorhabditis elegans. The first investigation uncovered that poly(A) selection skews sequencing results toward mRNAs with longer poly(A) tails, distorting our understanding of RNA populations and dynamics. This insight underscored the need for more inclusive sequencing approaches, which ultimately set the stage for applying Nanopore Sequencing to our core questions in NMD.The second focus centers on resolving the biochemical pathways by which animal cells attack NMD target mRNAs—an essential process for both cellular quality control and transcriptome regulation. Using single-molecule nanopore sequencing, we investigated the fates of NMD-targeted mRNAs, revealing that these targets undergo deadenylation and decapping at levels similar to normal mRNAs. We demonstrated that SMG-5, a protein previously implicated in deadenylation and decapping, is crucial for SMG-6-mediated endonucleolytic cleavage. Our results support a model in which NMD factors act in concert to degrade NMD targets in animals via an endonucleolytic cleavage near the stop codon, while deadenylation and decapping serve as routine aspects of typical mRNA (and NMD target mRNA) maturation and decay rather than exclusive features of NMD. Together, these studies contribute to refining RNA sequencing methodologies and deepening our understanding of NMD’s molecular mechanisms, offering insights that advance our knowledge of RNA biology.

Power flow (PF) analysis is critical to power system operation and planning. Nowadays, renewable energy power generation has been widely installed in power grids because they are environmentally friendly. The high penetration of renewable energy brings significant fluctuations to the power system states. Probabilistic power flow (PPF) analysis aims to characterize the probability properties of voltage phasors with stochastic power injections.

Exploiting the impressive capability of neural networks (NNs) in complex function approximation, we utilize the NN as a rapid PF solver in real-time applications. Motivated by residual learning, the first work proposes a new NN structure based on the physical characteristics of PF equations. Specifically, we add a linear layer between the input and the output to the multilayer perceptron (MLP) structure. We design three schemes to initialize the NN weights for the shortcut connection layer based on the linearized PF equations. Numerical results show that the proposed approach outperforms existing NN frameworks in estimation accuracy and training convergence. However, the branch flow estimation accuracy of the NN-based methods on some benchmark systems is lower than the linearized PF-based method. The inherent reason is that the NN outputs are voltage angles instead of voltage angle differences, while the latter determines the branch flows. To further improve the branch flow estimates, the second work separates the training of voltage magnitudes and phase angles due to their different properties. We incorporate the errors of voltage angle differences into the training loss function.

Based on PF equations, optimal power flow (OPF) analysis minimizes the total generation cost while subject to other operational constraints. To help the independent system operator (ISO) clear the real-time energy market, we develop an unsupervised learning-based framework to solve the OPF problem rapidly. We employ a modified augmented Lagrangian function as the training loss. The multipliers are updated dynamically during the training process based on the degree of constraint violation. Numerical results show that the dynamic updates of the penalty weight coefficient improve the feasibility of solutions compared to the fixed pre-assigned coefficient.

To ensure the PF balance, the NN predicts a subset of decision variables, and the remaining variables are obtained by a subsequent PF solver. However, the variable splitting scheme introduces heavy computation complexity when it comes to computing gradients in backpropagation. Hence, in the fourth work, we aim to reduce the total computational time of the NN to enable a daily update of the NN. We propose a physics-informed gradient estimation method based on a semi-supervised learning framework. We employ ridge regression to obtain pseudo-optimal solutions and build a hybrid dataset. We propose a batch-mean gradient estimation method based on the linearized Jacobian model to speed up the training process. Numerical results show that the proposed gradient estimation method achieves a similar convergence rate as the ground truth Jacobian. Moreover, the proposed method rapidly obtains near-optimal solutions, which is appealing in real-time applications.

This dissertation explores the feasibility of utilizing Silicon Carbide (SiC)Electroluminescence (EL) to estimate current from a SiC MOSFET’s body diode in classical power converter feedback control systems. The study delves into the current and temperature dependencies of SiC EL, demonstrating how light intensity at key wavelengths (390 nm and 500 nm) varies with current and temperature. By maintaining a constant junction temperature, the circuit’s electroluminescence is directly affected by a change in current, while a rise in junction temperature influences the light emission at different wavelengths. The work presents an experimental setup that integrates SiC EL with a closed-loop control system to regulate current in a buck converter. Results from the system demonstrate that SiC EL can be used to predict current, providing a basis for future motor drive torque regulation, speed control, and voltage control in power converters. The dissertation also addresses the challenges of low light intensity and nonlinearity in SiC EL measurements, proposing methods to optimize sensitivity and accuracy using avalanche photodetectors and calibration techniques. Despite limitations, such as the weak emission of SiC EL compared to direct bandgap materials, the research establishes a novel and effective approach for current estimation in power electronics applications, paving the way for improved control systems in power conversion and motor drives.

During protein synthesis, the ribosome must move the tRNAs and mRNA together in single codon steps after the addition of each amino acid to the polypeptide. This process of translocation is catalyzed by the GTPase elongation factor G in prokaryotes. EF-G hydrolyzes GTP during each round of translocation, yet the purpose of this energy expenditure is unclear. Here, we first ask how inhibiting GTP hydrolysis using GTP analogs and a mutant form of EF-G impacts the structural rearrangements of the ribosome that take place during translocation, monitored with Förster resonance energy transfer (FRET). We find that hydrolysis is required only for reverse rotation of the 30S head domain, an event that occurs late in translocation after the tRNAs and mRNA have completed the bulk of their movement. We then investigate the specific role and timing of phosphate (Pi) release from EF-G, which is delayed relative to hydrolysis and is the step responsible for the bulk of the energy derived from hydrolyzing GTP. We determine the timing of Pi release relative to the structural rearrangements of the ribosome by monitoring structural dynamics with FRET and in parallel observing the kinetics of Pi release with a fluorescence-based reporter. We find that Pi release occurs after forward head rotation and, surprisingly, is coupled to reverse intersubunit rotation. Further, we show that both Pi release and EF-G dissociation are required for reverse head rotation. To account for these findings, we propose that delayed Pi release prevents premature dissociation of EF-G; this ensures that the codon-anticodon duplex is stabilized by EF-G throughout its movement to prevent a frameshift. We conclude that the GTPase function of EF-G, rather than driving tRNA movement, is crucial for enforcing accuracy during translocation. This function may well extend to other translational GTPases such as IF2 and EF-Tu, which also exhibit delayed Pi release and have critical roles in enforcing accuracy during different steps of protein synthesis.

New tools are presented to generate simulated catalogs of microlensing events in theMilky Way from populations of primordial black holes tracing the dark matter halo. These Monte Carlo methods are orders of magnitude faster than the state-of-the-art simulations, and reduce the computational requirements from a large computer cluster to a laptop for full-sky surveys. A new statistic and method is demonstrated for highly efficient detection of microlensing events in multi-color star surveys. The background filtering is sufficiently strong to reject all lightcurves in a subset of NOIRLab Source Catalog data, while maintaining high efficiency on simulated events injected on the same data. These insights are combined to predict the exclusions on PBH dark matter the Legacy Survey of Space and Time will create over the ten-year survey.