http://www.rssboard.org/rss-specification 720 https://escholarship.org/uc/ucm_etd/rss Recent eScholarship items from UC Merced Electronic Theses and Dissertations Wed, 2 Sep 2026 16:54:26 +0000 https://escholarship.org/uc/item/9g29g10d The widespread adoption of electric vehicles (EVs) presents significant challenges in energy management, system safety, and large-scale fleet coordination. This dissertation develops and applies control and optimization strategies to solve critical EV problems at both component and fleet levels. A core theoretical contribution is the development of Recursive Control Barrier Functions (CBFs), a robust framework specifically designed for guaranteeing system safety under persistent constraints. This framework is then applied to the component-level problem of battery control, formulating fast charging and discharging as an optimal control problem to minimize charging and discharging time while strictly enforcing safety constraints on battery temperature, terminal voltage, and State of Charge (SoC). Separately, this work tackles the macro-scale challenge of emergency evacuation for EV fleets, which necessitates a focus on public safety and resilience. It develops a novel optimal control... https://escholarship.org/uc/item/9g29g10d Fri, 10 Jul 2026 00:00:00 +0000 Feng, Shuang https://escholarship.org/uc/item/9628p5ht Computing architectures have been evolving at a fast pace recently, driven by the emergence of large-scale applications such as artificial intelligence and the advance in hardware packaging technology. This trend calls for the birth of massively concurrent computing systems, driving the adoption of Multi-Chip-Module (MCM) GPUs and expansive multi-GPU clusters to deliver extreme computational capabilities. The conventional GPU system management methods are not adequate to handle the increasing concurrency that these scalable architectures bring forward, and will fail to deliver the full potential of the new hardware. Specifically, the massive parallelism of modern GPUs brings a substantial burden to shared memory management and network communication components, making the seamless orchestration of virtual addresses and inter-node data transfers an intricate challenge.The overall performance of highly concurrent GPU systems heavily depends on the operational efficiency of the... https://escholarship.org/uc/item/9628p5ht Fri, 10 Jul 2026 00:00:00 +0000 Feng, Yuan https://escholarship.org/uc/item/90x4r4fw Understanding the relationship between enzyme structure and function is a central goal in structural biology, with important implications for enzyme design, drug discovery, and biotechnology. In this dissertation, we investigate the role of structural flexibility and dynamics in enzyme catalysis using engineered enzymes as model systems and advanced X-ray crystallography techniques.Our results demonstrate that optimal catalytic function arises from a balance between rigidity and flexibility. In relatively simple, single-step reactions, increased rigidity can enhance catalysis by preorganizing the active site to closely match the transition state geometry, thereby reducing the need for conformational rearrangement. In contrast, more complex, multi-step reactions require enzymes to sample multiple conformational states to accommodate substrates, intermediates, and products throughout the catalytic cycle. In these cases, conformational flexibility becomes essential for efficient... https://escholarship.org/uc/item/90x4r4fw Fri, 10 Jul 2026 00:00:00 +0000 Seifinoferest, Behnoush https://escholarship.org/uc/item/8272q55r This dissertation examines fungal community structure in Sierra Nevada foothill oak woodlands by integrating multiple biological compartments and interaction types using ITS eDNA metabarcoding. This work adopts a multi-trophic framework in which fungal communities are shaped by host identity, environmental heterogeneity, and biotic interactions across spatial and temporal scales. Chapter 1 establishes this framework, positioning fungal communities as components of a connected ecological network spanning plant tissues, insect-generated microhabitats, and animal-mediated pathways. Chapter 2 shows that foliar endophyte communities in blue oak (Quercus douglasii) and interior live oak (Q. wislizeni) vary strongly by host species, site, and season, with blue oaks in exposed habitats harboring higher relative abundance of potential pathogens. Chapter 3 demonstrates that Cynipid galls act as localized ecological filters, hosting distinct, species-specific fungal assemblages while leaving... https://escholarship.org/uc/item/8272q55r Fri, 10 Jul 2026 00:00:00 +0000 Bivins, Christopher P. https://escholarship.org/uc/item/80r597g9 This dissertation investigates the impact of organic matter amendments (OMAs), with a focus on compost, on urban soil health and carbon storage. This research addresses key questions surrounding the impacts of OMAs not only in urban areas, but on soils generally, over month to decade timescales on soil biogeochemistry and management-relevant soil characteristics. Chapter two is a meta-analysis, focusing on how soil organic matter (SOM) and soil nutrients, such as nitrogen (N) and phosphorus (P), respond to different kinds of OMAs. We found that of the OMAs tracked – biochar, biosolids, and compost – biochar increased SOM% the most, by 6.5 percentage-points, while compost increased SOM% by 3.1. While biosolids saw the lowest SOM gains at 2.1, they had significantly greater gains in soil N and P. This chapter highlighted the urban OMA work done to this point, but stressed the importance of long-term and deep soil sampling, as most studies were 30cm or shallower and were monitored... https://escholarship.org/uc/item/80r597g9 Fri, 10 Jul 2026 00:00:00 +0000 Malone, Zachary David https://escholarship.org/uc/item/7s70c1v3 Immigrant victims of intimate partner violence (IPV) are at increased risk for barriers to and delays in health care. Local organizations can improve health outcomes for IPV victims by connecting them to essential services such as housing, counseling, legal services, and other safety-net programs. However, immigrants living in underserved areas are at disproportionate risk for barriers to health and safety net resources due to their limited eligibility for public benefit programs and these regions’ lack of infrastructure. The capacity and quality of organizations that serve immigrant IPV victims, such as legal aid and domestic violence agencies, and their variation in underserved areas, such as rural compared to urban regions, are largely unknown. I conducted a multi-method study to examine the immigration-related factors that influence health care access for immigrant IPV victims. First, I conducted 20 in-depth key informant interviews with legal aid and domestic violence organizations... https://escholarship.org/uc/item/7s70c1v3 Fri, 10 Jul 2026 00:00:00 +0000 Tafolla, Sharon https://escholarship.org/uc/item/7gb237nh Bilingualism and its effects on cognition have been extensively studied. However, many questions remain about how bilingualism relates to neurocognitive outcomes and the adequacy of assessments for capturing bilingual experience. This dissertation aimed to explore how variability in bilingual language use relates to attentional and error-related processing, and how current tools used to characterize language competence perform within diverse bilingual populations. Chapters 2, 3, and 4 used event-related potentials to test predictions of current models of bilingual effects on cognition. Chapter 2 used a visual search task to elicit the N2pc component as an index of spatial selective attention; bilingual experience did not predict N2pc amplitude or latency. Chapter 3 used a visual oddball paradigm to elicit the P3b component and found that greater bilingual language experience was associated with larger P3b amplitudes, suggesting more efficient allocation of attentional resources... https://escholarship.org/uc/item/7gb237nh Fri, 10 Jul 2026 00:00:00 +0000 Rodas De Leon, Nancy Elizabeth https://escholarship.org/uc/item/79336000 Soft corals in the Xeniidae family pulse in an energetically expensive manner to generate fluid flow, which benefits them biologically. This flow has been modeled using computationally intensive numerical methods, such as the Immersed Boundary Method, which limits the ability to explore a wide parameter space. To address this challenge, we introduce Gaussian Process Regression as a surrogate model to capture the kinematic and flow outcomes of these simulations. By training on a limited set of simulations, the surrogates enable accurate predictions across a broader range of parameters than those originally sampled. These surrogate models can then be optimized to identify the parameter combinations which maximize fluid flow. The resulting optimal parameters can further guide additional numerical simulations, creating an iterative framework for targeted exploration of the parameter space. This approach offers new insights into how the coral polyp musculature and local flow conditions... https://escholarship.org/uc/item/79336000 Fri, 10 Jul 2026 00:00:00 +0000 Malone, Sarah https://escholarship.org/uc/item/69v705tf Catalysis lies at the heart of modern chemical synthesis, underpinning the efficient production of pharmaceuticals, agrochemicals, and advanced materials. As the global scientific community increasingly prioritizes sustainability, there is a growing necessity to move away from catalytic systems that rely on precious metals such as iridium, rhodium, and palladium, elements that are not only scarce and expensive but also associated with significant environmental and supply-chain concerns. Homogeneous and organocatalytic approaches offer a compelling alternative for reducing the ecological footprint of chemical processes. In this context, Earth-Abundant transition metals and main-group elements have emerged as powerful platforms for sustainable catalysis. Despite their enormous potential, their widespread adoption remains hampered by several unresolved challenges, including susceptibility to side reactions, air- and moisture-sensitivity requirements, high catalyst loadings, and harsh... https://escholarship.org/uc/item/69v705tf Fri, 10 Jul 2026 00:00:00 +0000 Kaur, Harleen https://escholarship.org/uc/item/68p9w1dc In this dissertation, we investigate the photophysics involved in electronic excitation and photodetachment processes using single-determinant models developed within the delta-self-consistent field (∆-SCF) framework. Our results indicate that these single-determinant based methods in most cases yield excellent agreement with experimental data and perform competitively with modern multi-reference and response-based approaches.Chapter 1 provides a foundational background on the quantum mechanical principles that form the basis of modern quantum chemistry calculations. After establishing the core theoretical framework required for constructing and analyzing electronic wave functions, the two most prominent SCF methods in quantum chemistry, Hartree-Fock (HF) and Density Functional theory (DFT), are discussed in detail. The chapter then concludes with an overview of advanced correlated methods developed to describe excited states and photodetachment processes beyond a single determinant... https://escholarship.org/uc/item/68p9w1dc Fri, 10 Jul 2026 00:00:00 +0000 Bovill, Andrew Johnston https://escholarship.org/uc/item/5225b30h Relative camera pose estimation aims to estimate the camera movement between two images of the same environment. Traditional methods that are based on keypoint matching and recent methods that regress the camera pose with neural networks both show promising results on standard 3D datasets. However, multiple benchmark results show a large variation across image pairs, and it’s hard to determine a method that performs best on any image pair. Even the ’outdated’ method can produce the most accurate pose estimation under certain conditions. In this work, we present an agentic approach that leverages the profile of each image pair and uses a data-driven method to select the best estimator from existing methods without knowing the ground truth camera pose. We evaluate the workflow on multiple 3D datasets, and our method outperforms the current SOTA methods on multiple metrics. It provides a new solution to address the variation challenge of existing methods in different circumstances... https://escholarship.org/uc/item/5225b30h Fri, 10 Jul 2026 00:00:00 +0000 Du, Shangjie https://escholarship.org/uc/item/4z9228c6 This review paper synthesizes evidence across six domains —behavioral science, neuroscience, language, developmental psychology, computational modeling, and philosophy— to argue that sensorimotor information is organized by a single experiential principle: when sensorimotor systems are repeatedly co-active, they become coupled, so that re-presenting one element later reactivates its partners. This framework explains phenomena as diverse as classical conditioning, embodied cognition, cross-modal illusions, embodied word meaning, early empathy, expertise-driven cortical remapping, and the extended-mind thesis, and it provides a distributed, action-centered foundation for my forthcoming dissertation research. https://escholarship.org/uc/item/4z9228c6 Fri, 10 Jul 2026 00:00:00 +0000 Malaie, Soran https://escholarship.org/uc/item/4kk918c9 Growth mixture models (GMMs) constitute a versatile statistical framework for capturing population heterogeneity in longitudinal data by identifying latent classes of growth trajectories. Since their introduction in 1999, the models have been extended and disseminated through sustained methodological advancements, yet uncharted gaps remain in their implementation. This dissertation develops and evaluates advanced Bayesian methods to address two such gaps, predictor selection and missing data handling, across two studies. In Study 1, I focused on predictor selection in conditional GMMs. Beyond the usefulness of GMMs in describing heterogeneous growth patterns, predicting these patterns offers an enhanced understanding of substantive phenomena. However, the question of how best to select important predictors has remained unaddressed in the GMM literature. As a principled approach to this issue, I proposed Bayesian variable selection via shrinkage priors and evaluated seven priors... https://escholarship.org/uc/item/4kk918c9 Fri, 10 Jul 2026 00:00:00 +0000 Heo, Ihnwhi https://escholarship.org/uc/item/4gh572h1 Advanced computational modeling holds transformative potential, yet its impact is often limited by practical barriers: the complexity of numerical methods, the scarcity of high-quality data, and the need for specialized expertise. This dissertation introduces a cohesive body of work aimed at overcoming these challenges, delivering a framework of predictive tools unified by a central theme: accessibility. The core thesis is that by intelligently integrating numerical simulation and machine learning, we can create powerful models that are more straightforward to implement, robust with limited data, and intuitive for domain experts to use.Our work begins by addressing implementation and data accessibility. We first develop the Least-Square Discretization (LSQD) method, a novel meshless solver for elliptic PDEs designed to be accessible by removing the need for complex meshing. Concurrently, to make modeling accessible in data-scarce clinical environments, we construct a high-fidelity... https://escholarship.org/uc/item/4gh572h1 Fri, 10 Jul 2026 00:00:00 +0000 Terasaki, Gbocho Masato https://escholarship.org/uc/item/47t7g9ng Green hydrogen is widely recognized as a promising energy carrier that enables deep decarbonization and integrates variable renewable energy sources. However, its large-scale deployment remains constrained by high production costs and operational challenges associated with intermittent renewable electricity.This thesis presents an hourly techno-economic analysis of green hydrogen production in off-grid hybrid solar–wind systems directly coupled to electrolyzers, without battery storage or grid interaction. To determine the drivers of cost, the study focuses on the impact of renewable resource complementarity and electrolyzer sizing on hydrogen production cost. An hourly modeling framework (8760 h) was developed to simulate renewable electricity generation, electrolyzer operation, hydrogen production, and the Levelized Cost of Hydrogen (LCOH). Renewable generation profiles are obtained using the System Advisor Model (SAM), while a custom model evaluates electrolyzer performance... https://escholarship.org/uc/item/47t7g9ng Fri, 10 Jul 2026 00:00:00 +0000 Dotta Brugman, Maria Florencia https://escholarship.org/uc/item/43n8k3w4 Modern computing systems increasingly rely on heterogeneous architectures that integrate general-purpose processors with specialized accelerators such as GPUs and deep learning engines. To meet the performance and energy demands of modern workloads, these systems depend heavily on optimization techniques including speculative execution, transfer learning, hardware acceleration, and dynamic power management. While these mechanisms substantially improve efficiency, they also introduce new forms of shared state, resource contention, and externally observable behavior, thereby expanding the attack surface. As a result, security vulnerabilities in modern platforms often emerge not from a single component, but from interactions across multiple layers of the computing stack. This dissertation investigates such cross-layer vulnerabilities in heterogeneous computing systems and develops new insights into how performance-oriented design choices can unintentionally expose sensitive... https://escholarship.org/uc/item/43n8k3w4 Fri, 10 Jul 2026 00:00:00 +0000 Rafi, Mujahid Al https://escholarship.org/uc/item/3zz6k188 This dissertation comprises three complementary papers that examine the mechanisms through which mindfulness influences stress regulation and how these processes may differ for individuals whose regulatory systems have been shaped by adverse childhood experiences (ACEs). The first is a theoretical paper that introduces a novel framework, the Mindfulness Recalibration Model, which reconceptualizes mindfulness training as an experience-dependent process through which both somatic and cognitive-affective practices jointly recalibrate bottom-up physiological and top-down neurally mediated stress regulatory systems. The second is an experimental study that tests whether and how a general mindful state, cultivated through a brief induction, translates into enhanced coping appraisals when individuals subsequently encounter an unexpected acute stressor, and whether these effects differ as a function of ACE exposure. Results indicated no significant differences between the mindfulness... https://escholarship.org/uc/item/3zz6k188 Fri, 10 Jul 2026 00:00:00 +0000 Beylin, Natalie Rebecca https://escholarship.org/uc/item/3pf755ph Scientific discovery depends on our ability to model the physical world accurately --- and for many of the most consequential problems in biology and medicine, the geometry of the domain is not a detail to be approximated away but the very mechanism driving the phenomenon of interest. This dissertation develops and applies computational methods for partial differential equations on complex, irregular, and continuously deforming geometries, demonstrating that spatial resolution reveals biological insights that well-mixed or geometry-free approaches fundamentally cannot. The first contribution is the Least Squares Discretization (LSQD) method, a meshless framework for solving elliptic PDEs on scattered point distributions without requiring a mesh, a weak formulation, or quadrature rules. The method achieves h-P convergence across a broad range of geometries, boundary conditions, and problem configurations, remains robust in degenerate cases where traditional formulations may... https://escholarship.org/uc/item/3pf755ph Fri, 10 Jul 2026 00:00:00 +0000 Kucherova, Anna https://escholarship.org/uc/item/3291w8dc Rational design of molecules for chemical transformations and optoelectronic devices requires an understanding of excited state electronic transitions and the pathways connecting stable conformers on potential energy surfaces (PES). In this work, we demonstrate how computational methods like PES scans can be used to elucidate and design new reagents for chemical transformations. We also show how to use photophysical characterization of compounds, like charge transfer (CT), oscillator strengths, and absorption profiles can be used to design and chemically tune optical properties. We then show how Δ-SCF methods can be used to model CT states qualitatively using difference densities and quantitatively with CT distance metrics against methods like TD-DFT. Lastly, a new method for transition state (TS) optimization is developed and evaluated on complex topologies like the Müller-Brown and Schlegel surfaces. https://escholarship.org/uc/item/3291w8dc Fri, 10 Jul 2026 00:00:00 +0000 Sarabia, Cristian https://escholarship.org/uc/item/2kb0h3dk Individuals often face decisions in which the consequences of their actions for others are uncertain, and in many cases this uncertainty is actively maintained. A growing literature documents that individuals avoid information about the social consequences of their choices, a behavior commonly referred to as strategic ignorance. This dissertation argues that such ignorance is not merely an individual choice, but a socially embedded phenomenon shaped by interactions with others, internalized norms, and beliefs about the behavior of others. Using a series of experiments based on moral wiggle room environments, I examine how different features of the social context influence both the willingness to acquire information and the behavioral consequences of remaining ignorant. The results reveal a striking asymmetry: while many individuals avoid acquiring information themselves, others are readily willing to provide it. When information is imposed, a substantial share of decision-makers... https://escholarship.org/uc/item/2kb0h3dk Fri, 10 Jul 2026 00:00:00 +0000 Hua, Tony https://escholarship.org/uc/item/20n0m25k This dissertation develops a family of projection-based numerical methods for incompressible flow on adaptive, collocated grids. Here, projection methods act as the primary computational mechanism through which numerical stability is enforced, physical features are represented sharply, and the geometric structure of incompressible flow is respected. In a broader sense, these methods constitute a conceptual framework for modeling incompressible flows, enabling the development of more accurate, flexible, and physically consistent simulations of complex fluid phenomena. Chapter 2 develops a stable collocated adaptive mesh framework for the incompressible Navier-Stokes equations. While collocating variables simplifies algorithmic design and data structures, it disrupts the discrete representation of analytical operators that is typically automatic in staggered grid formulations. This challenge is addressed through the design of a stable nodal projection method tailored to node-based... https://escholarship.org/uc/item/20n0m25k Fri, 10 Jul 2026 00:00:00 +0000 Blomquist, Matthew https://escholarship.org/uc/item/0xt755fn The San Joaquin Valley is a rural, agricultural area in California with high rates of poverty and food and water insecurity. Applications of soil fumigants and pesticides in agriculture are widely understood to be a leading cause in groundwater contamination. From the 1940s until the mid-1980s in California, agricultural production often involved the use of pesticides, particularly those containing 1,2,3-Trichloropropane (TCP). After numerous studies found that TCP caused cancer and infertility in mice, TCP was banned from agricultural use in the 1980s. TCP, however, remains in groundwater at concentrations exceeding state regulations, especially in the San Joaquin Valley. Previous toxicology studies suggest that TCP is unsafe even at its lowest detection limit, 0.005 μg/L. One of the most assumed but understudied determining factors of contamination level is precipitation. This study statistically analyzes Groundwater Ambient Monitoring and Assessment (GAMA) Groundwater... https://escholarship.org/uc/item/0xt755fn Fri, 10 Jul 2026 00:00:00 +0000 Rosales, Angela A https://escholarship.org/uc/item/0x4830g7 Time-evolving systems arise throughout science and engineering whenever a governing equation is used to describe how a state evolves in time. A central goal in these settings is to accurately simulate, forecast, or infer this evolution, but practical progress is often limited by a combination of computational and data constraints. Common obstacles include expensive multiscale discretizations, temporal nonlocality induced by memory terms, and limited observations available for calibration. These difficulties are especially acute in quantum and electron dynamics, where relevant state spaces grow exponentially and near-term quantum hardware yields imperfect, noisy measurements. To this end, this dissertation develops equation-aware, data-driven methods that bring expensive, often intractable, computations within practical reach while preserving accuracy on targeted objectives. Specifically, the contributions are threefold.(1) For integro-differential equations (IDEs) motivated by... https://escholarship.org/uc/item/0x4830g7 Fri, 10 Jul 2026 00:00:00 +0000 Bassi, Hardeep https://escholarship.org/uc/item/7xg315gq Although shared decision-making is promoted as a cornerstone of patient-centered care, patients may sometimes need to educate their health-care providers to receive appropriate treatment. Across two studies (n Study 1 = 204; n Study 2 = 202), we examined the frequency and emotional impact of this experience. Study 1 used qualitative methods with a sample of convenience (college undergraduates) to explore participant experiences of this theorized construct. Study 2 used qualitative and quantitative methods with a sample of adults who are living with a chronic condition to further define the construct and examine associations with willingness to seek future care. Qualitative findings across both studies revealed that patients often needed to explain basic aspects of their condition, correct provider misconceptions, and advocate for standard or alternative treatments. This experience was described by a majority of participants in Study 1 (52.9%), and almost all participants (89.7%)... https://escholarship.org/uc/item/7xg315gq Tue, 10 Mar 2026 00:00:00 +0000 Coward, Charlie https://escholarship.org/uc/item/759270nn California’s agriculture, an over 60 billion dollar industry in 2024, faces significant challenges from two directions. Extreme temperatures and evaporation increase water requirements, while government policy, such as the Sustainable Groundwater Management Act, restricts the amount of water that can be pumped out of groundwater basins, limiting the amount of available water. As droughts in California are a repeating problem and state law continues to promote sustainability, solutions to optimize water use are increasingly important. One promising approach is agrivoltaics, which combines farming with solar panels to provide shade that reduces heat stress and water evaporation. While stationary agrivoltaic systems make it so the land is dual use, being used for farming and energy generation, they also restrict the use of large farm equipment, limiting the productivity and farming potential. To overcome these limitations, Simulation Based Sizing of an Off-Grid, Mobile, Agrivoltaic... https://escholarship.org/uc/item/759270nn Tue, 10 Mar 2026 00:00:00 +0000 Koerber, Andrew https://escholarship.org/uc/item/5xq3m8xs Despite the considerable attention given to host societies' attitudes toward immigrants, relatively little is known about immigrants' attitudes toward these societies. This study seeks to address the absence of research by examining levels of institutional confidence among immigrant and native-born Americans. Analysis of the 2022 General Social Survey reveals no statistically significant differences between native-born citizens and immigrants in baseline levels of institutional confidence. However, additional analysis shows that stratification beliefs shape immigrants’ institutional confidence in ways that do not extend to their native-born counterparts. I discuss the implications of the positive effect of holding higher stratification beliefs for immigrants possessing greater confidence and more favorable attitudes toward the United States, in contrast to native-born Americans, highlighting how the rhetoric of the American Dream and meritocracy still resonates strongly among... https://escholarship.org/uc/item/5xq3m8xs Tue, 10 Mar 2026 00:00:00 +0000 Zhao, Aaron https://escholarship.org/uc/item/5ww549b8 Telepresence robots enable remote interaction when physical presence is not possible, yet effective control remains a major challenge across desktop, mobile, and wearable settings. This dissertation designs, develops, and evaluates control systems for telepresence robots with a focus on performance, usability, safety, security, and user preference. We first introduce Tele Driver, a desktop interface inspired by driving simulators that maps steering wheel, pedals, and gear shifting to robot navigation. A user study shows it enables faster and more accurate operation than a keyboard-and-mouse interface and reduces perceived workload. We then present TiltWalker, a mobile interface that supports one-handed teleoperation using directional tilt gestures while preserving clear camera views. Iterative studies establish comfortable tilt ranges, appropriate control-display mappings, and demonstrate superior speed, accuracy, and user preference over a baseline mobile method. Next, we develop... https://escholarship.org/uc/item/5ww549b8 Tue, 10 Mar 2026 00:00:00 +0000 Zand, Ghazal https://escholarship.org/uc/item/2n12z3f8 Buildings account for over 40% of the nation’s energy consumption, with heating, ventilation, and air-conditioning (HVAC) systems responsible for nearly half of that use. Yet, many buildings still struggle to maintain occupant comfort efficiently due to limited sensing fidelity, the absence of real-time occupancy data, and non-adaptive control strategies. This dissertation presents a unified progression of data-driven sensing and control frameworks that advance building intelligence from accurate occupancy estimation to adaptive, energy-efficient HVAC control. First, we develop MODES, a Multi-sensor Occupancy Data-driven Estimation System that fuses vibration and thermal signals through a data-driven optimization and particle filtering pipeline. MODES improves occupancy estimation accuracy by up to 40% and enables HVAC systems to achieve up to 77% energy savings and 10% comfort improvement through occupancy-informed control. Building upon this foundation, we introduce TODOS, a... https://escholarship.org/uc/item/2n12z3f8 Tue, 10 Mar 2026 00:00:00 +0000 Rajabi, Hamid https://escholarship.org/uc/item/0rd0503h Mercury (Hg) is a globally distributed contaminant of critical environmental and public health concern. Although naturally occurring, its prevalence and mobility in the environment have been greatly amplified by industrial activity, leading to widespread contamination of aquatic and terrestrial ecosystems. Understanding the biogeochemical processes that control Hg speciation, transformation, and mobility is essential for predicting its fate and mitigating its ecological impacts. This dissertation investigates the coupled roles of redox conditions, mineral associations, and ligand interactions in regulating Hg cycling across multiple environmental contexts, ranging from managed wetlands to contaminated soils and sediments.The first component of this work examined Hg transformation and transport within seasonally managed wetlands to evaluate how hydrological manipulation and redox dynamics influence Hg speciation and methylation potential. Over three years of field sampling, wetland... https://escholarship.org/uc/item/0rd0503h Tue, 10 Mar 2026 00:00:00 +0000 Jones, Danielle Jeanine https://escholarship.org/uc/item/06m341f4 Metal halide perovskites (MHPs) are at the forefront of materials research due to their superior optical and electronic properties, which have led to their utilization in a wide range of applications. Despite their performance, their susceptibility to environmental degradation remains a critical challenge for widespread deployment. This dissertation investigates the response of MHPs to various environmental factors and explores novel functionalities that can be enabled by leveraging these interactions. This work first explores the surface passivation of MHPs using perovskite quantum dots (PQDs). Different coverage densities of PQDs, varying in composition and surface functionalization were deposited onto methylammonium lead iodide (MAPI) thin films. We found that low-density PQD coverage passivated surface defects of MAPI, enhancing emission intensity and recombination lifetimes. An investigation into space-based applications revealed the resilience of MAPI films to ultraviolet... https://escholarship.org/uc/item/06m341f4 Tue, 10 Mar 2026 00:00:00 +0000 Arteaga, Jorge https://escholarship.org/uc/item/8r18r01g As the global energy transition accelerates, from coal and natural gas to renewable sources like solar and wind, ensuring year-round grid reliability becomes increasingly complex due to the intermittency of these renewables. This dissertation investigates two key strategies for addressing these challenges: (1) deploying energy storage to shift excess generation to periods of need, and (2) leveraging flexible electric vehicle (EV) charging to better align demand with renewable supply. Using capacity expansion modeling, the study explores how these solutions can enhance both the cost-efficiency and resilience of renewable-powered grids. Results show that promoting midday charging of light-duty EVs can significantly reduce grid reliance on costly energy storage, saving billions of dollars in California's renewable-energy-driven grid. To address the high costs of non-residential charging infrastructure, the study proposes widespread deployment of low-cost level-1 chargers at workplaces,... https://escholarship.org/uc/item/8r18r01g Mon, 9 Mar 2026 00:00:00 +0000 ZareAfifi, Farzan https://escholarship.org/uc/item/7b62p1bv Household food waste is a major contributor to global greenhouse gas emissions and a critical barrier to achieving state and national climate goals. Efforts to divert food waste from landfills as a means to reduce methane emissions requires local data. In California’s San Joaquin Valley, an agricultural region with limited composting infrastructure and scarce household-level data, understanding food waste generation is essential for designing effective organics diversion strategies. This study quantified household food waste generation, examined socioeconomic drivers of waste, and assessed willingness to participate in different composting systems across three San Joaquin Valley counties. We used a citizen science approach that recruited 134 households to sort and record food waste daily for seven consecutive days. These data were coupled with demographic and behavioral surveys. We found an average per-capita food waste generation of 76.5 ± 6.3 kg y-1, consistent with global estimates... https://escholarship.org/uc/item/7b62p1bv Mon, 9 Mar 2026 00:00:00 +0000 Wilson, Aaryn https://escholarship.org/uc/item/6st1v55k Multi-object tracking (MOT) in dynamic and crowded environments is still a difficult problem in computer vision. The main reasons are that objects often get occluded and many of them look very similar. Current tracking-by-detection methods, such as StrongSORT, rely a lot on appearance-based re-identification (Re-ID) models to keep track of object identities. However, features extracted by traditional Convolutional Neural Networks (CNNs) are not very robust. When real-world issues like motion blur, partial occlusion, or low-light conditions appear, these features often break down and cannot provide stable identity information. This weakness leads to more identity switches (IDSW) and fragmented trajectories, which directly limit the performance of the tracker in complex scenes. In this thesis, I propose a new MOT framework to address this problem by replacing the usual CNN feature extractor with a stronger embedding from a pre-trained Denoising Diffusion Probabilistic Model... https://escholarship.org/uc/item/6st1v55k Mon, 9 Mar 2026 00:00:00 +0000 Hu, Yiqi https://escholarship.org/uc/item/6d30t78g Background: Mindfulness-based interventions are found to effectively reduce stress and improve mental health outcomes. Yet, it is not always clear how the mindfulness skills of attention and acceptance that are expected to produce these improvements develop throughout the intervention. This knowledge gap is especially pertinent for novice meditators learning these skills for the first time, including whether some individuals are more prone to learning them. Purpose: Using a randomized waitlist-controlled trial, we tested the effect of the mindfulness app Headspace on changes in attention and acceptance over eight weeks among participants new to mindfulness meditation. Further, we tested the moderating effects of trait mindfulness and perceived stress. Method: Non-faculty university employees were randomized to a Headspace or waitlist control condition. Trait mindfulness and perceived stress were measured at baseline. Ecological momentary assessment survey data for attention and... https://escholarship.org/uc/item/6d30t78g Mon, 9 Mar 2026 00:00:00 +0000 Peña, Mercedes https://escholarship.org/uc/item/110486x1 Battery-powered mobile robots are increasingly deployed in orchards for tasks such as sensing, monitoring, and crop management. In these settings, robots must traverse long distances on unstructured terrain with limited opportunities to recharge, so energy usage directly constrains mission duration and coverage. However, accurately predicting power consumption in agricultural environments is challenging because it depends on a combination of route geometry, speed profiles, terrain conditions, and payload configuration. This thesis investigates how these factors influence the power demand of an agricultural mobile robot and develops a context-aware learning framework for forecasting its power consumption. The study uses a real-world dataset collected from 72 autonomous navigation trials executed on three representative route types: rectangular paths, capsule-shaped circuits, and unstructured trajectories in a commercial pistachio orchard. Each trial includes synchronized motor... https://escholarship.org/uc/item/110486x1 Mon, 9 Mar 2026 00:00:00 +0000 Abedi, Ali https://escholarship.org/uc/item/9mf734m4 In this thesis, we experimentally studied the microtubule-based active nematic confined in a microfluidic channel. By using controlled channel geometries, we examine the influences of confinement on flow behavior, defect dynamics, and motion. We used Fluorescence microscopy, particle image velocimetry (PIV), and manual defect tracking to characterize the system’s flow and dynamic behavior. PIV measurements reveal the complex flow fields generated by ATP-driven active stresses in the system and capture spontaneous flow and velocity fluctuations associated with the defect motion and active turbulence, while remaining statistically steady over long times. The root-mean-square velocity vrms provides a quantitative measure of the speed of the flow field in both channels, the bounded active flow fluctuations indicating a nonequilibrium steady state. The analytical data show that the left channel shows slightly higher velocities and larger fluctuations, demonstrating the sensitivity... https://escholarship.org/uc/item/9mf734m4 Sat, 7 Mar 2026 00:00:00 +0000 Shahida, Shahirin https://escholarship.org/uc/item/7kk6r0fn Cluster states are a class of highly entangled quantum states which act as an important resource in measurement-based quantum computing. The generation of cluster states has been studied in many platforms which typically depend on high-fidelity, controlled quantum logic gates performed between the qubits. These qubit logic gates can be difficult to implement in many practical physical systems. Here, we propose a scheme to generate cluster states of spin defects in a phononic quantum network. Such a system lacks the direct coupling which is typically required for generating cluster states. In our scheme, the qubits are coupled to mechanical modes in an optomechanical system and the interaction between these mechanical modes and a common optical cavity facilitates the entanglement generation. The interaction between the qubits and the mechanical mode is tuned such that the overall state of them both exhibits two-level system (TLS) behavior. Thus, their combined systems interacting... https://escholarship.org/uc/item/7kk6r0fn Sat, 7 Mar 2026 00:00:00 +0000 Mazzei, Mitch https://escholarship.org/uc/item/37t6p66m A growing body of research shows how administrative burdens for receiving social benefits act as rationing policies that reduce access to resources through enacting learning, compliance, or psychological costs on those seeking resources. Burdens are commonly enacted through policies, paperwork, and procedural steps that beneficiaries go through to gain access to resources. Parallel research has shown the growth of inequalities from the increasing financialization of social benefits through their management according to financial logics or the delivery as loan-based benefits. This paper integrates these recent theoretical advances regarding inequalities in social and educational programs by theorizing institutional student debts as financialized administrative burdens. The paper then elaborates the theory by evaluating whether institutional student debts vary across public higher education institutions in relationship to their interrelated governance arrangements, marketization,... https://escholarship.org/uc/item/37t6p66m Sat, 7 Mar 2026 00:00:00 +0000 Martin, Andrew https://escholarship.org/uc/item/2g36t276 Sketches are data structures that can probabilistically estimate various statistics. One such statistic is cardinality — the count of elements in a set. In database systems, the cardinality of a query operation — the size of its result — predicts the computational cost of subsequent operations and is therefore crucial for optimizing query execution. However, a sketch is typically only applicable to the limited set of queries for which it was specifically constructed. For example, a sketch for the cardinality of a join query may be inapplicable to another query with the same join but different filter conditions. Although highly accurate, their lack of generality hinders their adoption for improving systems. This dissertation explores the use of machine learning models for generating sketches to estimate the cardinality of arbitrary queries on the fly. Experiments show that the generated approximations are accurate, while being both more time and memory-efficient than the original... https://escholarship.org/uc/item/2g36t276 Sat, 7 Mar 2026 00:00:00 +0000 Tsan, Brian https://escholarship.org/uc/item/8v2188f6 Biochar is a carbon-rich material produced through the pyrolysis of agricultural and forestry residues, playing a critical role in managing excess organic waste in California’s Central Valley and Sierra Nevada regions. Beyond waste valorization, biochar’s structural and chemical tunability make it a promising platform for environmental remediation, catalysis, and energy storage applications. This dissertation examines the mechanistic interactions between non-thermal plasma and biochar and explores how plasma-based post-treatments can enhance biochar properties by modifying its structure, surface chemistry, and electrical behavior at substantially lower temperatures than conventional activation processes (800–1000 ◦C). A key focus is understanding how biochar interacts with non-thermal plasma during exposure, and whether these interactions can be leveraged to enhance the inherent properties of the material. Two complementary studies were conducted to address this. The first utilized... https://escholarship.org/uc/item/8v2188f6 Fri, 6 Mar 2026 00:00:00 +0000 Gomez, Hector https://escholarship.org/uc/item/8n91w06x Viral infections continue to pose significant threats to global health, even in the age of modern medicine. These pathogens manipulate host cellular processes by reprogramming metabolic pathways to support their replication. Understanding how viruses hijack host metabolism is crucial for elucidating their mechanisms of pathogenesis and identifying potential therapeutic targets. Metabolomics, an emerging field within systems biology, enables comprehensive analysis of cellular metabolites, providing a direct reflection of the phenotypic alterations that occur during infection. This dissertation examines the dynamic relationship between viral infection and host metabolic reprogramming, emphasizing the role of metabolomics in uncovering viral strategies that disrupt key metabolic pathways. Chapter I traces the current understanding of host–viral interactions, with a focus on metabolism. I discuss how the application of metabolomics has advanced knowledge of viral manipulation of host... https://escholarship.org/uc/item/8n91w06x Fri, 6 Mar 2026 00:00:00 +0000 Sanchez, Bailey-J Christopher https://escholarship.org/uc/item/7mr5j298 Recent advancements in quantum sensing have pushed measurement sensitivities far beyond classical limits. These advancements in measurement capability have improved our ability to probe quantum systems, producing datasets that were previously inaccessible or prohibitively noisy. This access to high quality data has shifted quantum science research from measurability to interpretability.The observable, measured data from quantum systems is frequently not the quantum system’s parameters of interest, but rather a complicated function of these parameters. Recovering the underlying physical parameters from the measured data constitutes an inverse problem. We consider a variety of problems of this form where the quantum systems are modeled by Hamiltonian matrices, the measured data corresponds to the eigenenergies of these Hamiltonians, and the parameters of interest are contained within the Hamiltonian matrices. By measuring the system’s excited energy levels at varying electric fields,... https://escholarship.org/uc/item/7mr5j298 Fri, 6 Mar 2026 00:00:00 +0000 Wright, Kyle https://escholarship.org/uc/item/6sv3z253 Hydrogels and elastomers are commonly used substrates for mechanobiology and cell mechanics studies to mimic the mechanical properties of native extracellular matrices (ECM). This is partly due to the ability to tune the elastic properties within the range relevant to soft tissues (Young's modulus of 1–40 kPa) by adjusting their chemical composition with relative simplicity. For affine gels, two (of three) elastic constants -- the Young’s modulus (E), the shear modulus (G), and the Poisson’s ratio (ν) -- are needed to describe the purely elastic response to external forces. While the need for accurate estimation of the Young’s modulus is well recognized, the importance of correctly estimating the Poisson’s ratio is often overlooked. For instance, for computational ease, in traction force microscopy measurements, hydrogels are approximated as incompressible (ν = 0.5). Here, I present evidence indicating that the Poisson’s ratio of soft materials is directly proportional to the... https://escholarship.org/uc/item/6sv3z253 Fri, 6 Mar 2026 00:00:00 +0000 Smith, Ariell https://escholarship.org/uc/item/6pg9v68z Cilia are evolutionarily conserved organelles that play critical roles in brain development, yet their spatiotemporal dynamics in the embryonic and postnatal ventricular system remain poorly characterized. Here, we investigate the morphological transitions of primary and motile cilia on the surface of the developing brain ventricles, focusing on radial glial cells (RGCs) and their differentiation into ependymal (E1/E2) cells and adult neural stem (B1) cells. Using whole-mount staining and a cilia-specific transgenic mouse model, we reveal region- and time-specific patterns of cilia remodeling. During embryogenesis, we observed a progressive increase in cilia length and a significant decrease in cilia density in both the neocortex and ganglionic eminences, with a sharp elongation of cilia in the ventral telencephalon (LGE and MGE) at E17, coinciding with early multiciliogenesis. In contrast, the lateral wall undergoes postnatal multiciliogenesis, with a rapid transition from single-ciliated... https://escholarship.org/uc/item/6pg9v68z Fri, 6 Mar 2026 00:00:00 +0000 Torres Gutierrez, Oscar Julian https://escholarship.org/uc/item/69c970r0 California’s agricultural systems face dual pressures of water scarcity and climate variability intensified by the Sustainable Groundwater Management Act (SGMA). This dissertation integrates spatial modeling, climate analysis, and agronomic feasibility assessments to evaluate three high-value specialty crops — coffee, olives, and saffron — as models for sustainable diversification in the San Joaquin Valley and other water-limited regions. Using high-resolution historical and future climate data, thermal suitability thresholds were mapped and coupled with floodplain and irrigation demand models to define zones of opportunity and constraint. The results show that coffee can be cultivated in coastal and foothill microclimates, olives provide a water-smart alternative to tree nuts, and saffron offers ultra-low water use and seasonal dormancy aligned with SGMA objectives. Collectively, these findings demonstrate how data-driven crop substitution and system optimization can reduce water... https://escholarship.org/uc/item/69c970r0 Fri, 6 Mar 2026 00:00:00 +0000 Prewitt, Kenneth Burrell https://escholarship.org/uc/item/47x879jh This dissertation investigates how nanomaterials—nanoparticles and amphiphilic additives —modify phase behavior and structural organization in two classes of liquid crystal systems: a thermotropic nematic host doped with functionalized quantum dots and a lyotropic lipid membrane that models the pulmonary surfactant. In the thermotropic study, calamitic organic ligands were designed to compatibilize CdSe/ZnS quantum dots with 5CB, enabling controlled phase-transition–templated assembly into architectures such as microshells and filamentous networks. Early-stage nematic domain formation and the accompanying nanoparticle transport were acquired with high-speed fluorescence microscopy and quantified with ilastik-based machine-learning segmentation, allowing the influence of cooling rate and nanoparticle concentration on nucleation density, initial domain growth, and quantum dot redistribution at the moving phase boundary to be determined. In the lyotropic system, Langmuir-trough... https://escholarship.org/uc/item/47x879jh Fri, 6 Mar 2026 00:00:00 +0000 Ochoa, Jocelyn https://escholarship.org/uc/item/40x7r91j Social support is crucial for effective management of type 1 diabetes (T1D), but support is not always helpful and sometimes can undermine management efforts. Aside from family and friends, adults with T1D can utilize the diabetes online community (DOC) to receive support from other people with diabetes. Support from experientially similar others is theorized to be different and preferred to support from close significant others, but this has not previously been explored in the context of T1D. We assessed social support from family/friends and from experientially similar others on the DOC to examine: a) differences in satisfaction with support and frequency of multiple forms of social support (i.e., emotional and instrumental support, controlling and illness avoiding behaviors) between sources, b) whether source moderates associations between support and indicators of diabetes management (i.e., self-efficacy, self-management behaviors, diabetes distress, and HbA1c), and c) whether... https://escholarship.org/uc/item/40x7r91j Fri, 6 Mar 2026 00:00:00 +0000 Ellis, Emily Brooke https://escholarship.org/uc/item/3pp8c22j Biomass conversion plays an important role in producing renewable fuels and reducing the environmental burden of waste organic residues. The work in this dissertation focuses on the thermal conversion of biomass through pyrolysis and combines physics-based modeling, controlled experiments, and machine learning to improve predictability and process control. A one-dimensional finite volume model was developed to describe the propagation of pyrolysis-front inside a top-lit updraft (TLUD) pyrolyzer. The model incorporates detailed heat transfer, gas flow, species transport, and reaction kinetics. It was used to study the influence of operating conditions such as air velocity, porosity, and particle properties on temperature evolution, front speed, and energy release. A machine learning model was then used to predict pyrolysis front propagation and peak solid temperature in a multi-feedstock experiment using peach and plum pits, utilizing a digital twin framework for fuel flexibility,... https://escholarship.org/uc/item/3pp8c22j Fri, 6 Mar 2026 00:00:00 +0000 Nasef, Ziad https://escholarship.org/uc/item/38d5n1n5 Filipino Americans cross race, class, and nativity hierarchies. Because of a colonial relationship between the United States and the Philippines, agrarian and working class Philippine nationals had freer passage to the US mainland, growing their US population much later into the 20th century than for other Asian origin groups. Upon the resumption of migration, immigration policy after 1965 created pan-ethnic Asian incentives for middle- and professional class individuals. Using US Census data, I show that US born Filipinos whose families arrived before 1965 have persistently lower rates of college attainment than those whose families arrive after 1965. Drawing on 34 in depth interviews with Filipinx Americans, I show that cultural embeddedness, cultural identity, and shifting racialized cultural status hierarchies for Filipinx may contribute to this divergence. My research contributes to empirical knowledge of Filipino Americans, a large but understudied group of Asian Americans. https://escholarship.org/uc/item/38d5n1n5 Fri, 6 Mar 2026 00:00:00 +0000 Colond, Jay Zeus https://escholarship.org/uc/item/30q8v3nc As the world shifts toward renewable energy, the major challenge of how to store energyfrom intermittent sources like wind and solar remains. Advanced batteries and other energy storage technologies offer promising solutions, but current manufacturing methods limit how effective and efficient these devices can be. This dissertation explores how 3D printed polymer electrode templates can help create improved energy storage devices. By printing thick, precisely designed electrode architectures, it's possible to increase the amount energy-storing material and improve performance. However, turning printed parts into electrically conductive components often requires pyrolysis at very high temperatures, which can cause parts to shrink or warp and produce little usable carbon. Chapter 1 of this dissertation surveys the existing literature on the additive manufacturing of pyrolyzed carbon from polymeric precursors. The background of the pyrolytic conversion of polymers to carbon is described... https://escholarship.org/uc/item/30q8v3nc Fri, 6 Mar 2026 00:00:00 +0000 Chiovoloni, Samuel https://escholarship.org/uc/item/1n0034f5 This dissertation investigates data-efficient methods for applications in image processing under conditions of sparse, limited, and noisy observations. Many imaging applications, such as synthetic aperture radar (SAR), non-destructive testing, and photon-limited systems, operate with low-dimensional measurements or limited labeled data. Deep learning approaches are typically known to require large datasets, and for large-scale optimization approaches, the problem is often ill-conditioned in nature. The goal of this research is to develop and analyze both learning-based and model-based frameworks that overcome these concerns in various image reconstruction applications despite these data constraints.The first part of the dissertation focuses on learning-based and deep learning strategies. A recurrent neural network (RNN) architecture is introduced for image denoising, emphasizing the impact of convolution padding on preserving spatial structure. This is followed by a study of machine... https://escholarship.org/uc/item/1n0034f5 Fri, 6 Mar 2026 00:00:00 +0000 Alvarez, Jacqueline https://escholarship.org/uc/item/18m7451c Under climate change, increasing drought stress has been directly linked to tree death in forests worldwide and indirectly linked to tree death via insect or pathogen outbreaks. Selecting tree seedlings with higher drought tolerance for reforestation may increase forest stability and carbon storage. On one hand, genome-environment associations (GEAs) can be used to predict seedling stress response by associating single nucleotide polymorphism (SNP) alleles with trees’ historical environment. On the other hand, symbiotic microbes can influence the growth and survival of plants under drought and could be used alone or in combination with genetic selection to increase seedling drought tolerance. However, few studies have investigated how the interaction of genotype and microbiome influences drought response in trees. Conifers represent ecologically and economically important tree taxa in many arid temperate regions. The goal of this study is to better understand conifers’ drought... https://escholarship.org/uc/item/18m7451c Fri, 6 Mar 2026 00:00:00 +0000 Du, Yinyue https://escholarship.org/uc/item/9334w08h Image restoration, particularly low-light and compression artifact removal, remains a challenging problem due to severe loss of illumination, color distortion, and high-frequency detail degradation. Traditional methods often struggle to preserve semantic content, while existing learning-based approaches require large annotated datasets and are limited in handling extreme underexposure. In this paper, we propose a novel reference-based image enhancement framework, ReGIE, that integrates Retinex priors, reference image conditioning, and a diffusion model guided via ControlNet. Our method leverages Retinex-based decomposition of the degraded image to provide structural illumination priors and conditions the generative process using a semantically similar reference image retrieved from a multi-view dataset. To mitigate the over-smoothing effects common in latent diffusion models, we introduce a high-frequency fidelity constraint using the Discrete Fourier Transform and Sobel filtering,... https://escholarship.org/uc/item/9334w08h Mon, 13 Oct 2025 00:00:00 +0000 Seth, Siddharth https://escholarship.org/uc/item/5r80408c Spliceosomes are ancient cellular machines found across eukarya that perform essential work in the maturation of pre-mRNAs. Falling into two classes: the U2 (major) and the U12 (minor), these spliceosomes detect and remove their respective classes of spliceosomal introns from mRNA molecules using a set of particles consisting of snRNA molecules and an assortment of proteins and are guided to the introns based on sequence motifs. Though spliceosomes and their associated introns are found in nearly all lineages of eukaryotes, most of our knowledge of these features come from the study of a select few model organisms. Further, while the spliceosome is thought to be highly conserved between lineages, there are several well-documented examples of independent transformations in diverse lineages. Our understanding of the pan-eukaryotic spliceosome is severely lacking, and the nature of spliceosomal transformation has not been fully explored. This dissertation aims to elucidate the... https://escholarship.org/uc/item/5r80408c Fri, 10 Oct 2025 00:00:00 +0000 Bowser, Bradley Alan https://escholarship.org/uc/item/462545ss Artificial Intelligence (AI) language models, such as GPT-mini (ChatGPT) or OpenAi-o1, have demonstrated exceptional capabilities in language processing, logical reasoning, and symbolic tasks. However, spatial reasoning, which encompasses the ability to mentally manipulate objects, understand spatial relationships, and predict structured spatial sequences, remains largely uncharted in the context of large language models (LLMs). Spatial reasoning is fundamental to human cognition and relies on specialized neural structures such as cortical maps, as suggested by Dehaene et al. Unlike humans, LLMs lack explicit spatial architectures, raising important questions about their ability to perform tasks that require spatial transformations, recursion, and geometric compression.This study investigates the extent to which OpenAi gpt-4o or GPT-mini can perform spatial reasoning tasks by adapting Dehaene’s geometric language framework into a purely verbal input-output paradigm. Tasks include... https://escholarship.org/uc/item/462545ss Fri, 10 Oct 2025 00:00:00 +0000 Ingale, Ajinkya https://escholarship.org/uc/item/2tz1642m Formation flight refers to aircraft flying in a coordinated pattern while maintaining a close proximity to each other. This configuration is frequently used in public service operations, such as military aviation. The military uses different configurations depending on the mission requirements. For the purposes of this report aerial refueling was researched. Aerial refueling is a formation flight configuration that consists of a tanker and a receiving aircraft. While in flight, a refueling mechanism physically connects the tanker and receiving aircraft. Possessing this capability yields economic and strategic advantages. Although the concept of aerial refueling has existed for nearly a century, it remains a complex and demanding procedure for modern aircraft and pilots. Therefore, using OpenFOAM to compute 2D simulations of this procedure can provide insight on how the Cl and Cd values are impacted by the receiver being in the wake. With the data collected it indicates that Cl... https://escholarship.org/uc/item/2tz1642m Fri, 10 Oct 2025 00:00:00 +0000 Chavez Jr, Rafael https://escholarship.org/uc/item/0x2137hb Tiered memory architecture, which integrates both fast memory (such as DRAM) and slow memory (such as non-volatile memory), has emerged as a promising and cost-effective solution to address the growing need for large memory capacity in a wide range of modern and data-intensive applications. While the use of tiered memory allows for increased capacity at reduced cost, fast memory remains a significant contributor to overall system expense. Therefore, optimizing and potentially reducing the size of fast memory is critical for improving memory utilization and lowering production costs in practice. However, deciding on the appropriate fast memory size is a challenging problem due to the complex interplay between workload characteristics, application sensitivity to memory performance, and the overhead associated with page migration between memory tiers.This thesis presents Tuna, a novel system designed to determine the optimal mal fast memory size in tiered memory environments by modeling... https://escholarship.org/uc/item/0x2137hb Fri, 10 Oct 2025 00:00:00 +0000 Chen, Shangye https://escholarship.org/uc/item/9vc2h62h Decision forests, also known as tree ensembles, have become an important model class in machine learning, particularly for tabular data, due to their excellent predictive accuracy, robustness, and minimal requirements for hyperparameter tuning. Remarkably, their widespread success has been achieved despite traditional training methods being largely heuristic, neither the individual trees nor the forest as a whole explicitly optimize a clearly defined loss function. Typically, these ensembles rely on methods such as bagging and boosting, where base learners are generated through greedy recursive partitioning. This dissertation proposes an alternative, optimization-driven framework for learning decision forests. Building upon the recently introduced Tree Alternating Optimization (TAO) algorithm, we first show significant improvements when explicitly optimized base learners replace the standard, heuristically constructed trees within two of the most popular boosting algorithms: AdaBoost... https://escholarship.org/uc/item/9vc2h62h Tue, 7 Oct 2025 00:00:00 +0000 Gabidolla, Magzhan https://escholarship.org/uc/item/9ft8s0gk Bilingualism is an extensively studied phenomenon in language research and has been demonstrated to have a considerable impact on how individuals process linguistic information. Research indicates that individuals who speak a second language exhibit greater variability in their language performance compared to monolinguals, potentially attributable to disparities in their linguistic background across various bilingual communities. In the current study, we explored how second-language exposure and proficiency modulate the effects of surprisal on sentence processing during natural reading. The results show surprisal is a strong predictor of reading time. However, the study did not find any significant effects on reading time associated with second language proficiency or exposure. https://escholarship.org/uc/item/9ft8s0gk Tue, 7 Oct 2025 00:00:00 +0000 Liu, Chen https://escholarship.org/uc/item/8xj3m233 Global agriculture faces mounting pressure to reduce greenhouse gas emissions while maintaining productivity and enhancing soil health. Livestock operations worldwide contribute 32% of global methane emissions, with dairy systems representing a significant portion through manure management, while synthetic fertilizer dependence continues to degrade soils and contaminate water systems across diverse agricultural regions. California exemplifies these global challenges: its dairy industry produces 40 million tons of manure annually, accounting for 25% of the state's methane emissions and creating urgent pressure to meet the 40% emission reduction of 2013 levels mandated under Senate Bill 1383 by 2030. Simultaneously, sustainable and regenerative agriculture initiatives encourage sustainable alternatives to synthetic fertilizers that enhance soil health while maintaining productivity. This dissertation investigates biochar co-compost—created by adding biochar at the beginning of the... https://escholarship.org/uc/item/8xj3m233 Tue, 7 Oct 2025 00:00:00 +0000 Gonzales, Melinda L https://escholarship.org/uc/item/89w9k95z Quantum mechanics describes the properties of nature both at and below the atomic scale. Quantum chemistry calculations have rapidly expanded into all areas of chemistry, underlining the importance of understanding modern methods of electronic structure calculation. Density functional theory is among the several prevailing approaches to calculating molecular properties, and it is arguably the most popular method due to its efficient balance between computational cost and accuracy. In this method, one does not attempt to calculate the many-body wave function, but instead calculates the electronic probability density and the corresponding molecular electronic energy. Originally developed for ground-state systems, density functional theory can be extended to finite-temperature systems via the thermal density functional theory framework, which is commonly used in high-energy-density physics to simulate matter under extreme conditions. However, there are currently only a few approximate... https://escholarship.org/uc/item/89w9k95z Tue, 7 Oct 2025 00:00:00 +0000 Harding, Brittany Penelope https://escholarship.org/uc/item/843185xw Emotion understanding relies on children’s emotion scripts, which include facial and vocal expressions, labels, eliciting events, and behavioral consequences related to specific emotions. Particularly important for social functioning, though profoundly understudied, is children’s understanding of behavioral consequences of emotions; to modify their own behaviors, children need to understand what an angry, sad, etc. person is likely to do. This study examined 3- to -6-year-old children’s (N = 20; M = 53.4 months old) understanding of behavioral consequences for four emotions (anger, disgust, sadness, fear). Children saw narrated stories of a cartoon character who encountered a specific emotion eliciting event. Children then choose the behavioral consequence (e.g., hiding in response to fear) that completed the story from two images. Analyses indicated that children reliably chose the correct behavioral consequence for the disgust (75%; p = .04) and anger conditions (79%; p = .02).... https://escholarship.org/uc/item/843185xw Tue, 7 Oct 2025 00:00:00 +0000 Özden, Zeynep B. https://escholarship.org/uc/item/7mn100jg This dissertation investigates the acoustic radiation forces arising from wave scattering between closely-situated sound-hard spheres, with direct implications for the phenomenon of acoustic binding. Motivated by experimental findings in the Kleckner Lab at UC Merced, we develop a numerical and analytical framework to compute surface fields and resulting forces with high precision, even in the challenging near-field regime. Using boundary integral equations (BIEs) and spherical harmonic expansions, we resolve the scattering problem for two spheres in regimes where multiple scattering and coupling effects are non-negligible. We begin by establishing a boundary integral formulation that allows direct computation of surface fields, circumventing the need for full domain discretization. We then expand the field using a spherical harmonic basis and derive a Galerkin discretization scheme capable of capturing both independent and coupled scattering behaviors. A key contribution is the... https://escholarship.org/uc/item/7mn100jg Tue, 7 Oct 2025 00:00:00 +0000 McCullough, Cory Storm https://escholarship.org/uc/item/7ck2r9r1 Understanding light–matter interactions in plasmonic nanostructures and metamaterials is crucial for a wide range of applications from biochemical sensing and high resolution imaging to optical cloaking. This leads us to study Wave scattering by negative materials, which in turn leads to a Helmholtz equation with sign-changing coefficients, a regime in which classical coercivity of the variational formulation breaks down at resonance frequencies. This thesis develops and analyzes Galerkin methods to address the theoretical and computational challenges of this near-resonant plasmonic scattering problem. On the theoretical side, we characterize the near-resonance behavior by constructing the resonances and associated resonant modes for canonical geometries (e.g. a penetrable ball). As the material parameter ϵm approaches a critical value, the solution exhibits unbounded growth in its amplitude, with a loss of coercivity and can admit, for some wavenumbers, a plasmonic resonance.... https://escholarship.org/uc/item/7ck2r9r1 Tue, 7 Oct 2025 00:00:00 +0000 Latham, Benjamin Jarred https://escholarship.org/uc/item/6ww807q0 This article demonstrates how international PhD students at UC Merced, a newly designated R1 university in rural California, navigate and construct a sense of belonging. Based on thirty semi-structured interviews with students from Asia, the Middle East, Africa, and South America, the study extends frameworks of belonging and academic integration, applying graduate student development theory to contextualize these dynamics. Belonging was found to be largely selfcreated, shaped by language barriers, advisor relationships, peer networks, and misaligned resources. The findings suggest that belonging is essential to academic persistence and closely tied to students’ ability to engage with the university, often through personal resilience and informal networks shaped by varying levels of academic and cultural capital. Institutional responsiveness—or its absence—significantly shapes these experiences. This study contributes to student development scholarship by showing how international... https://escholarship.org/uc/item/6ww807q0 Tue, 7 Oct 2025 00:00:00 +0000 Haghverdian, Camellia https://escholarship.org/uc/item/6n8547xx Content creation aims to empower users to visualize complex ideas and generate rich media content with minimal manual effort. Recent advancements in generative models have significantly automated the creative process, producing increasingly realistic and high-quality outputs. These developments lower the barrier for non-professional users to create compelling artistic content without requiring extensive domain expertise.In particular, techniques such as Neural Radiance Fields (NeRF), 3D Gaussian Splatting, and recent text-to-video generation models have shown great promise in facilitating 3D content creation and synthesizing high-fidelity videos from natural language prompts. Despite these impressive results, several key challenges remain. Existing models often struggle with: 1) maintaining temporal and spatial consistency across multiple views and video frames; 2) offering fine-grained controllability over individual subjects; and 3) generating complex content involving articulated... https://escholarship.org/uc/item/6n8547xx Tue, 7 Oct 2025 00:00:00 +0000 Huang, Hsin-Ping https://escholarship.org/uc/item/5wm8t88k As California transitions toward a 100% renewable electricity grid, addressing the seasonal variability of renewable resources becomes crucial to ensuring reliability and cost-effectiveness. While power outages and rolling blackouts have historically occurred during the summer in a fossil-fuel-dominated grid, this dissertation demonstrates that in a renewables-driven grid, the critical period for resource adequacy shifts to the winter months due to reduced solar and wind generation.Several pathways exist to mitigate this seasonal challenge, including expanded transmission capacity, demand-side flexibility, long-duration energy storage (LDES) adoption, and resource diversification. This dissertation focuses on strategic approaches to integrating solar and wind energy while minimizing the reliance on LDES to manage seasonal mismatches efficiently.Using RESOLVE as the capacity expansion modeling tool, this study evaluates the effectiveness of south-facing solar mounting configurations... https://escholarship.org/uc/item/5wm8t88k Tue, 7 Oct 2025 00:00:00 +0000 Mahmud, Zabir https://escholarship.org/uc/item/5rq126k3 The emergence of class stratification within the populations of the Northern Frontier of Mesoamerica has long intrigued researchers, sparking ongoing study and vigorous debates. The prevailing argument has been that this region was characterized by an egalitarian hunter-gatherer lifestyle or simpler settled chiefdoms with some Mesoamerican cultural characteristics. Here, I, adopt Household Archaeology and Processual Archaeology approaches to delve deeper into the social dynamics of Cerro de en Medio (CDEM), an archaeological site from the Epiclassic period (600-900 CE) located in the Mexican state of Aguascalientes. With its 215 structures, including identified domestic architecture, CDEM provides an exceptional opportunity to investigate social organization. Through archaeological excavations of three domestic compounds, chemical analysis of occupation floors, and comprehensive study of archaeological materials, this project aims to uncover the intricacies of production activities... https://escholarship.org/uc/item/5rq126k3 Tue, 7 Oct 2025 00:00:00 +0000 Campos Martinez, Miriam Selene https://escholarship.org/uc/item/5m20w147 This dissertation examines how transracial adoptee artists use visual art to challenge dominant narratives of adoption, reclaim epistemic authority, and produce embodied forms of knowledge that resist erasure and redefine adoptee experience. Drawing on interdisciplinary frameworks from critical adoption studies, critical race theory, postcolonial theory, and visual culture studies, this project argues that adoptee-created art is not merely expressive, but theoretical, political, and epistemological. These artists disrupt tropes of rescue, assimilation, and gratitude by foregrounding complexity, rupture, and relational care. Based on interviews and visual analysis of the work of twenty transracial adoptee artists, the study explores how their creative practices construct counter-visualities that intervene in cultural discourse and assert alternative modes of knowing. The chapters trace a progression from the historical development of critical adoption studies, to close readings... https://escholarship.org/uc/item/5m20w147 Tue, 7 Oct 2025 00:00:00 +0000 Lecy, Nicolette Lin Zhigao https://escholarship.org/uc/item/54s5h219 Machine learning and data science are extremely powerful tools for prediction, analysis, system identification, and estimation. This dissertation presents two new methods for system identification using neural networks. The first method, covered in Chapter 2, uses neural shape functions to learn vector fields from data. The second, presented in Chapter 5, integrates neural networks into a continuous-time Markov chain framework to learn propensity functions from data. This integration allows for propensity functions that depend on outside covariates. In addition to proposing these methods, this dissertation also examines fairness in machine learning algorithms and proposes preprocessing methods to increase fairness and equity in machine learning algorithm outputs. These methods, Separate and Equity resampling, are introduced in Chapter 3 in the context of predicting suicide death. The Equity resampling method is further examined in Chapter 4. The overarching goal of this dissertation... https://escholarship.org/uc/item/54s5h219 Tue, 7 Oct 2025 00:00:00 +0000 Reeves, Majerle Elizabeth https://escholarship.org/uc/item/4pv6b1k3 Giant vesicles are micrometer scale, self-assembling, spherical closed lipid bilayers that mimic biological cells. They have commonly been used as models to study biophysical properties of cellular membranes. Advances in the bottom-up approach of synthetic biology have extended the applications of giant vesicles to be used as scaffolds for the construction of protocells as well as their potential as microreactors. Giant vesicles as microreactors carrying out different biochemical reactions like gene expression and nucleic acid replication have predominantly been synthesized using water-in-oil emulsion techniques. Here, I present the use of a novel thin-film hydration technique of giant vesicle assembly named ‘One-Stepped Modulated PAPYRUS’ (OSM-PAPYRUS) to conduct polymerase chain reaction (PCR) within vesicles. I used a commercially available PCR mix called MangoMix and a standard PCR mix for the reactions. I found that in-vesicle reaction is possible with both mixtures, but... https://escholarship.org/uc/item/4pv6b1k3 Tue, 7 Oct 2025 00:00:00 +0000 Khurram, Nimra https://escholarship.org/uc/item/4bz6q9bn This work investigates the fundamental physical principles of self-assembly and structure of systems of liquid crystals and nanoparticles, lipids, and polymers from biological samples. By understanding the fundamental mechanisms behind the assembly processes involved in these systems, and by studying the properties of the resultant structures, we can find ways to apply this knowledge to the development of new materials for a variety of applications. I report on studies of the growth of nematic liquid crystal (LC) domains in the presence of nanoparticles; investigating which experimental parameters will affect the spontaneous growth of an ordered nematic phase from a disordered isotropic phase upon cooling. When the nanoparticles are surface modified with liquid crystal-like ligands, this process has been shown to create hollow microstructures with different morphologies. I report on the creation of these microstructures using newly synthesized ligands with varying length and electron... https://escholarship.org/uc/item/4bz6q9bn Tue, 7 Oct 2025 00:00:00 +0000 Wheeler, Alauna C https://escholarship.org/uc/item/41f5h0zp The ethnoracial liminality of Filipinx students may predispose them to microaggression exposure at higher rates than other Asian groups. Using a large survey of early college undergraduates, I ask if Filipinx students report more frequent racial microaggressions than other Asian groups. I investigate this by running OLS regression on student exposure to peer, staff, and faculty microaggressions, controlling for gender, socio-economic status, and nativity. I find, with some limitations, that Filipinx students report higher rates than Asian ‘honorary white’ groups, confirming their collective black placement. I also find overall that peers are a more frequent source of racial microaggressions than staff or faculty. I suggest further investigation of Filipinx status in interaction with socio-economic status, migration and multiracial status. https://escholarship.org/uc/item/41f5h0zp Tue, 7 Oct 2025 00:00:00 +0000 Colond, Jay https://escholarship.org/uc/item/4121p3xx Chronic pain is one of the most common reasons that adults seek medical attention. Individuals’ treatment and illness beliefs may influence their behaviors and health outcomes. We aimed to test a process model, based on the common sense model of illness self-regulation, specifying relations among illness and treatment beliefs, and health outcomes in individuals experiencing chronic pain using synthesized data from multiple studies. Studies (k=93, N= 18,262) reporting relations between model constructs were identified in a systematic database search. We tested the fit of our model with the pooled matrix of correlations among the variables from a multi-level meta-analysis using meta-analytic structural equation modeling. We also tested the moderating effects of having a chronic pain illness diagnosis on these relationships. The analyses revealed non-zero averaged correlations among the relations between the illness (control, timeline, identity, consequences, cause, coherence, and... https://escholarship.org/uc/item/4121p3xx Tue, 7 Oct 2025 00:00:00 +0000 McKinley, Lauren Elizabeth https://escholarship.org/uc/item/28k911m9 Objective: The theory of planned behavior is a prominent social cognition theory widely applied to predict physical activity. While prior meta-analyses of applications of the theory in this behavioral context have provided generalized support for its effects, limitations and evidence gaps have been noted such as syntheses of evidence on the unique effects of differentiated theory constructs, and effects of indirectly-measured belief-based constructs, on intentions and behavior. Further, there is a dearth of research on moderating conditions under which theory effects vary. To address these limitations and evidence gaps, we used synthesized data from a large-sample meta-analysis of applications of the theory in physical activity contexts to test effects of the basic theory and, importantly, particularly effects of differentiated and indirectly-measured theory constructs on physical activity intentions and behavior as well as examining effects of conceptually- and methodologically-salient... https://escholarship.org/uc/item/28k911m9 Tue, 7 Oct 2025 00:00:00 +0000 Simpson-Rojas, Danielle Victoria https://escholarship.org/uc/item/9tr31798 In this work, the effects of i) assisting compounds, ii) substrate, iii) polymer-grafted lipids, and iv) lipid nominal surface concentration on giant unilamellar vesicle (GUV) assembly are explored. GUVs are micron-sized, closed bilayers that are assembled in aqueous environments from amphiphilic molecules. There is interest in using GUVs for cell models and drug delivery applications. GUVs have been utilized to study various biophysical phenomena, including cytoskeletal mechanics, lipid phase separation, and membrane protein function. There is also interest in using GUVs as microreactors and for drug delivery. Despite the numerous methods developed for assembling GUVs, the mechanisms of assembly are currently unknown. Here, using quantitative methods developed in the Subramaniam lab, the size distributions and molar yields of GUV assembly are reported. Mechanisms for assembly are also proposed based on the quantitative data collected. The effect of assisting polymers templated... https://escholarship.org/uc/item/9tr31798 Mon, 6 Oct 2025 00:00:00 +0000 Cooper, Alexis https://escholarship.org/uc/item/9p45184g The processes resulting in biodiversity have been a focus of evolutionary biology since the field’s inception. Historically, vicariance, natural selection, and niche evolution have been considered primary generators of lineage diversity at macroevolutionary levels while other processes such as demography, sexual selection, and sexual conflict have been relegated to the microevolutionary scale. However, theoretical advancements coupled with empirical studies have increasingly supported the importance of demography, sexual selection, and sexual conflict facilitated by environmental heterogeneity in generating diversity at a broader evolutionary scale. This dissertation examines divergence between and among lineages at three distinct evolutionary scales to understand the roles of sexual processes in generating and maintaining divergence. We generated a reference genome for the common side-blotched lizard (Uta stansburiana) and used it to assemble a transcriptomic dataset comparing... https://escholarship.org/uc/item/9p45184g Mon, 6 Oct 2025 00:00:00 +0000 Fellows, Samuel R https://escholarship.org/uc/item/9h06s4pf The primary and overarching goal of this dissertation is to disambiguate consciousness through a combined approach of physiological experimentation and philosophical investigation. I start with a behavioral experiment in Chapter 2 on inattentional deafness that lays the groundwork for a follow-up EEG paradigm. Both experiments were conducted using a 2 × 2 design that manipulated perceptual and/or cognitive load while measuring auditory awareness. Our behavioral results aligned with load theory. We found that high perceptual load reliably led to failures in sound detection, while effects of cognitive load were more nuanced. In Chapter 3, I extended this paradigm to explore the neural correlates of consciousness (NCCs) via EEG analysis, with a central focus on how task relevance and the act of reporting may be conflated with perceptual awareness. The neurophysiological results of the experiment included both event related potentials (ERPs) and oscillatory activity. For ERPs, I found... https://escholarship.org/uc/item/9h06s4pf Mon, 6 Oct 2025 00:00:00 +0000 PROULX, ANDREW MEHDI https://escholarship.org/uc/item/98x0p82n This project investigates the responsiveness of federal courts to civic action, as it focuses on whether and how protest activity and high-profile police killings influence judicial decision-making in excessive force cases within the U.S. Courts of Appeals. Across an array of national- and circuit-level analyses, this study finds limited evidence that protest activity directly sways judicial votes. Instead, the research uncovers a significant negative correlation between high-profile police killings of African Americans and the likelihood of a pro-law enforcement vote. In other words, high-profile killings of African Americans such as those of Michael Brown or Tamir Rice seem to decrease the likelihood of a vote in favor of a law enforcement officer accused of excessive force. Federal judges appear to react to these salient triggering events, but not to the protest movements that follow. The findings also reveal the enduring influence of judicial ideology, which remains the most... https://escholarship.org/uc/item/98x0p82n Mon, 6 Oct 2025 00:00:00 +0000 Robles, Ricardo https://escholarship.org/uc/item/8tx8s6zw Internet of Things (IoT) applications are conventionally classified into small-scale and large-scale deployments. Small-scale applications, such as smart homes, typically span 0–100 meters and utilize short-range communication technologies like Zigbee and Bluetooth Low Energy (BLE). In contrast, large-scale applications, exemplified by Microsoft’s FarmBeats, extend across 1–5 kilometers and rely on long-range radios such as LoRa and SigFox. Recently, a new and distinct class of deployments, termed mesoscale IoT applications, has emerged. These operate over intermediate distances (approximately 0.1 to 1 kilometer) and often repurpose either short-range or long-range radios beyond their optimal design ranges. Mesoscale applications face unique challenges, including unpredictable link quality, lack of dedicated radio infrastructure, and environmental variability, which are not adequately addressed by existing protocols designed for small- or large-scale systems. To address this gap,... https://escholarship.org/uc/item/8tx8s6zw Mon, 6 Oct 2025 00:00:00 +0000 Shanmuga Sundaram, Jothi Prasanna https://escholarship.org/uc/item/7mm391sj Familial Hypercholesterolemia (FH) is an inherited disorder leading to increased risk of premature cardiovascular disease. This risk can be ameliorated through adherence to pharmacological treatment and salient lifestyle behaviors: physical activity participation and following a cholesterol-lowing diet. Identifying theory-based, modifiable determinants of these behaviors may inform development of interventions to promote participation in FH self-management behaviors. We aimed to identify the belief-based social cognition constructs likely to be associated with intentions to perform, and actual participation in, FH self-management behaviors in the extant research. We conducted a systematic database search to identify studies (k=10, N=1505) reporting relations between constructs from social cognition theories and intention toward, or actual participation, in self-management behaviors in FH patients. Relations among the constructs and FH selfmanagement behavioral intentions and past... https://escholarship.org/uc/item/7mm391sj Mon, 6 Oct 2025 00:00:00 +0000 Majeed, Rabia https://escholarship.org/uc/item/7jm0f45b Humans have a natural predisposition towards music– we sing, we dance, and we pay hundreds of dollars to see an artist perform live music. Despite these observations, no scientist today could answer with certainty why humans produce or enjoy music. One approach to understand musicality, or the biological foundations that support why we create, perform, and listen to music, is to examine how individuals who acquire a musical background, typically by practicing an instrument for several years, undergo neural changes that affect how they process complex sounds, such as music and speech. This dissertation outlines three experiments that span auditory perception and cognitive processing of auditory patterns, illusions, and adaptation with an emphasis on understanding how musical expertise shapes music and speech processing. The culmination of this work revealed that musicians, relative to nonmusicians, recruit higher-level resources to process music and speech. Together, this work... https://escholarship.org/uc/item/7jm0f45b Mon, 6 Oct 2025 00:00:00 +0000 Santoyo, Alejandra https://escholarship.org/uc/item/79m2f541 Skill acquisition is a fundamental aspect of human learning, underlying everything from motor control to social behavior. This dissertation investigates skill learning using three complementary approaches: behavioral analysis, electroencephalography (EEG), and a novel conceptual framework we call “abstract cognitive graphs.” In the first study, participants played a simple two-dimensional aiming game while their motor performance was tracked over time. Participants’ performance improved and behavioral variability decreased with practice following a power law trajectory, which aligns with classic models of skill acquisition. In the second study, we examined EEG data during an expanded version of the game to assess how neural activity changes with learning. We found that ERP increased during motor execution, while theta/alpha spectral power decreased around feedback evaluation. This suggests a shift toward reduced cognitive effort and more efficient neural processing. In the third... https://escholarship.org/uc/item/79m2f541 Mon, 6 Oct 2025 00:00:00 +0000 Kim, Liza https://escholarship.org/uc/item/78s4z1tn As bifactor modeling has become increasingly popular, concerns have been raised for its potential overuse due to its enhanced flexibility in fitting data. This issue is particularly pressing given the common but flawed practice of reusing the same data to generate and test structural hypotheses, as traditionally done for model selection. Another issue of bifactor modeling lies in its strong assumptions, such as the orthogonality of factors and the exclusion of cross-loadings. This study addresses both issues by using the Bayesian framework to estimate a confirmatory bifactor model that allows several cross-loadings. We argue that Bayesian estimation, with its flexibility and ability to incorporate prior knowledge, is a promising approach for estimating complex bifactor models with cross-loadings. We assess parameter recovery of such models under varying magnitudes of a general factor, different levels of cross-loadings, sample sizes, and prior specifications. Of particular interest... https://escholarship.org/uc/item/78s4z1tn Mon, 6 Oct 2025 00:00:00 +0000 Ivanov, Aleksandr https://escholarship.org/uc/item/75f243bx Earth system models exhibit order-of-magnitude disagreements in soil carbon projections, representing the largest uncertainty in climate predictions. This dissertation reveals a fundamental yet overlooked source of this uncertainty: the tyranny of small scales in biogeochemical modeling. Through three interconnected studies spanning microseconds to continental scales, we demonstrate how seemingly minor short-term processes create systematic biases that cascade through temporal and spatial scales to fundamentally compromise global carbon cycle predictions. Using continental-scale-analysis of hundreds of AmeriFlux stations, we first reveal pervasive systematic bias in soil respiration models arising from the universal practice of using average environmental conditions. Jensen’s inequality mathematically predicts these biases for nonlinear biogeochemical functions, with dry and shallow sensors exhibiting near 100% underestimation when using mean conditions. These systematic patterns... https://escholarship.org/uc/item/75f243bx Mon, 6 Oct 2025 00:00:00 +0000 Perez Rojas, Yulissa Tonatzin https://escholarship.org/uc/item/74b3g2nk Wildland-urban-interface fires account for billions in damage and loss of property and life. Structures in these fires are typically ignited by direct flame contact and/or embers being produced by already ignited vegetation and structures nearby. By preventing or delaying the ignition of these structures, we can provide time for evacuations or defensive measures to be put into place. This thesis aims to determine the ignition criteria and behavior for a novel wood-waste-derived material. The wood-composite material uses low-energy processing methods and is a recyclable material with the potential as a low-cost and sustainable housing component. A biodegradable and recyclable gelatin based flame-retardant is also applied to the wood-composite material samples. The objectives are to determine the wood composite ignition behavior properties and compare its performance with pinewood. As well as determining any ignition behavior improvements the flame retardant may provide the wood-composite,... https://escholarship.org/uc/item/74b3g2nk Mon, 6 Oct 2025 00:00:00 +0000 Avalos, Joseph https://escholarship.org/uc/item/71x9v54q Electronic structure theory calculations tell us about matter and its interactions at a quantum level. One of the most prevalent of these methods is density functional theory, which operates by mapping a set of interacting electrons onto a set of non-interacting ones. This is possible because the exchange-correlation energy functional approximation accounts for the non-mean interactions of electrons. Building on density functional theory’s foundation, many other methods have been developed such as ensemble or time-dependent density functional theory. The method of finite-temperature density functional theory is the most important method in this work. A common struggle among all of these methods is the concept of strong correlation. Within DFT, one way to interpret this issue is by considering the relation of potential and kinetic correlation energies. This is further complicated at finite temperature with the inclusion of entropic correlation and explicit and implicit temperature... https://escholarship.org/uc/item/71x9v54q Mon, 6 Oct 2025 00:00:00 +0000 Martinetto, Vincent https://escholarship.org/uc/item/6p58d1mk This dissertation investigates the spatial and social foundations of economic inequality through three empirical studies centered on geography, kinship, and community context. Each chapter explores a distinct but complementary mechanism that shapes long-run economic and demographic outcomes.The first chapter examines the enduring influence of historical state centrality on modern population distribution and political geography. Using a novel dataset of over 500 civilizations between 1000 and 1500 CE, I show that proximity to historical political centers predicts both historical and contemporary population density and the location of modern national capitals. These findings highlight the persistent demographic and political imprint of ancient civilizational hubs.The second chapter turns to the family as a key social institution shaping life trajectories. Using administrative data on birth and marriage records in Texas, I analyze how sibling sex composition affects marriage outcomes.... https://escholarship.org/uc/item/6p58d1mk Mon, 6 Oct 2025 00:00:00 +0000 Kim, Raymond Shinkwang https://escholarship.org/uc/item/6nh964pg What does it mean to take a ’4EA conceptualization’ of cognition? Why does variability matter? How do these ideas apply to interpersonal communication? What do virtual agents afford us? and finally, how do we experimentally test this framework? These are the questions that this dissertation seeks to answer and provide evidence for their importance. In chapter 1, the consummate concepts of 4EA cognition will be explicated and related to communication. In doing so, the relevance of variability will be argued, and the potential utility of virtual agents will be proposed. Chapter 2 represents a first effort combining all these lines of inquiry with an experiment where virtual agent stimuli vary on variability in specific ways. Subsequently, Chapter 3 is an extension, refining the manipulation of the virtual agent, and introducing listener eye-tracking. The introduction of this measure opens exploratory avenues for relating movement and gaze in complexity analyses. Chapter 4 shifts... https://escholarship.org/uc/item/6nh964pg Mon, 6 Oct 2025 00:00:00 +0000 Nguyen, Benjamin Hoang https://escholarship.org/uc/item/6hm349sp The optimal number of latent classes to capture heterogeneity in the population, known as class enumeration, has traditionally been approached through criterion-based and model-fit indices in both frequentist (Akaike, 1974; Celeux et al., 2006; Schwarz, 1978) and Bayesian frameworks (Asparouhov & Muthén, 2021; Heo et al., 2024). However, traditional approaches face significant limitations, such as the necessity for repeated pairwise comparisons (Asparouhov & Muthén, 2021; Depaoli, 2013; Heo et al., 2024; Liu & Song, 2018; van de Schoot et al., 2018). The Reversible Jump Markov Chain Monte Carlo (RJMCMC) method presents an alternative, data-driven approach that estimates the number of latent classes and model parameters simultaneously, thereby avoiding the need for model comparison. Previous work by Liu and Song (2018) demonstrated the application of RJMCMC within the mixture Structural Equation Modeling (SEM) framework, yet omitted factors like sample size, class proportions,... https://escholarship.org/uc/item/6hm349sp Mon, 6 Oct 2025 00:00:00 +0000 Jauregui, Madelin Gwineth https://escholarship.org/uc/item/62v1m44k This dissertation explores how structural modulation across molecular and mesoscale platforms governs material properties by examining two complementary systems: ligand-functionalized quantum dot assemblies and photoresponsive organic emitters. Despite differences in mechanism and scale, both systems demonstrate how externally tunable structural features can be harnessed to achieve functional outcomes, offering routes toward the rational design of responsive optical materials. This work addresses fundamental challenges in materials science by establishing predictive structure–property relationships that connect molecular-level design to macroscopic function.The first part focuses on colloidal CdSe/ZnS quantum dots (4.8 nm core diameter) functionalized with bis(imino)pyridine (BIP) ligands bearing different substituents (Me, H, IPr). Following ligand exchange under inert conditions, the functionalized QDs are dispersed in the thermotropic liquid crystal 5CB and self-assemble into... https://escholarship.org/uc/item/62v1m44k Mon, 6 Oct 2025 00:00:00 +0000 Karappilly Rajan, Arya https://escholarship.org/uc/item/5x06d3fs A vertical PV fence solar canal design is found to be competitive with canopy systems in terms of generation and water shading when built along the banks of N-S oriented canals with width < 25ft (7.6 m), particularly in areas with high soiling rates. A row of vertical PV along a single-bank remains an option under scenarios where canal accessibility is limited to a single side, making canopy systems impossible. The addition of a low mounted reflective tarp can improve total vertical system generation by 19%-24% (or 41%-44% with a hybrid reflector with an aluminized mylar top layer), though prolonged soiling on a reflector cleaned only by rainfall could lead to average lifetime values as low as 12% unless the material has non-stick properties to promote windblown cleaning. The presence of a reflective tarp also improves shading and reduces mass transit due to wind, resulting in an average 61±4% reduction in evaporation loss compared to an open canal. However, the 4-corner mounting... https://escholarship.org/uc/item/5x06d3fs Mon, 6 Oct 2025 00:00:00 +0000 Reagan, Jeremiah Bristow https://escholarship.org/uc/item/5s89j697 Both e-cigarettes and marijuana are growing in popularity and this phenomenon is even more pronounced among those who smoke cigarettes. Very little is known about the scale and consequences of e-cigarette and/or marijuana co-use among smokers and it is unclear whether the use of these products reduces or increases smokers’ exposure to harm over time. We hypothesized the following: 1) e-cigarettes will be primarily used for cigarette cessation and marijuana will be used for recreation; however, the most popular products will have potentially harmful components similar to cigarettes (e.g. nicotine in e-liquids and combustible marijuana products); 2) co-use will be associated with greater cigarette consumption and nicotine addiction indicators, 3) co-users will be more motivated to quit smoking cigarettes and engage in more cessation behaviors, and 4) current and recently quit smokers who use e-cigarettes and marijuana will report independent addictions to these products and recently... https://escholarship.org/uc/item/5s89j697 Mon, 6 Oct 2025 00:00:00 +0000 Halliday, Deanna Marie https://escholarship.org/uc/item/5m47645x Imaging hypoxia heterogeneity or low oxygenation distribution at high spatial resolution in deep tissues, such as bone marrow, is essential for understanding its role in stem cell proliferation and differentiation. Currently, although many imaging techniques can visualize oxygen distribution, they are limited either by high spatial resolution or by their ability to image oxygenation distribution in deep tissues. X-ray Luminescence Computed Tomography (XLCT) is a novel functional imaging modality that overcomes these limitations by achieving high resolution at the micrometer scale in tissues as deep as several centimeters. This is feasible because XLCT combines the advantages of X-ray imaging's high spatial resolution and optical imaging's high sensitivity. Furthermore, XLCT imaging has been introduced as a powerful new hybrid molecular imaging modality, capable of high-resolution imaging of deeply embedded x-ray excitable contrast agents in three dimensions (3D). This dissertation... https://escholarship.org/uc/item/5m47645x Mon, 6 Oct 2025 00:00:00 +0000 Zhang, Yibing https://escholarship.org/uc/item/5jz2p327 The present study introduced an integrated theoretical model linking constructs from achievement goal theory (AGT) and self-determination theory (SDT) with behavioral outcomes in physical activity, sport, and performance contexts. In the model, autonomous and controlled forms of motivation from SDT were proposed as mediators of associations between achievement goal orientations from AGT and behavioral outcomes. Model predictions were tested using meta-analytically synthesized data from available studies in the extant research literature. In addition, model effects were expected to vary according to salient methodological and contextual conditions tested in a series of moderator analyses of the meta-analytic model tests. A systematic literature search identified eligible studies reporting associations among AGT, SDT constructs, and behavioral outcomes (k = 66). The proposed model was fit to the meta-analytically pooled matrix of correlations among model constructs and behavioral... https://escholarship.org/uc/item/5jz2p327 Mon, 6 Oct 2025 00:00:00 +0000 Benoit, Thais Camargos Zanatta https://escholarship.org/uc/item/5fm8s02t The proliferation of energy storage, energy harvesting, and electrochemical devices drives demand for power sources with commensurate form factors. An electrochemical cell, such as a fuel cell or electrolysis cell, can be a solution for future energy sustainability.A key component in an electrochemical cell is the gas diffusion layer (GDL), a porous conductive layer that modulates the relevant transport of gas, liquid, compression load, heat, and electricity. Optimizing the GDL material and structure for efficient transport is critical to achieve high performance. Current baseline GDL materials are made of carbon fibers, which are difficult to alter the geometry, porosity, and pore size distribution.The fundamental purpose of this paper is to apply additive manufacturing techniques to create a 3-D microporous structure of the GDL using in-house-designed conductive ink. In a fuel cell, it is essential for the porous media to have optimized transport properties as well as scalable... https://escholarship.org/uc/item/5fm8s02t Mon, 6 Oct 2025 00:00:00 +0000 Sagar, Pulkit https://escholarship.org/uc/item/5c90q6xq Solid oxide fuel cells (SOFCs) are one of the most promising types of fuel cells, which have been extensively researched as a highly efficient generator of electrical energy from chemical resources. The high operating temperature of SOFCs (conventionally > 800°C) facilitates the kinetics of electrochemical reactions and charge transport through the cell, making the resulting power output viable for use even with metal oxide-based electrodes. However, a wide deployment of SOFCs has been significantly hampered by high costs, lengthy startup and shutdown times, thermal stresses, and unintended chemical reactions between components in SOFC's harsh operating environment. Lowering the working temperature to a more moderate range (600–800 °C) is a key to avoiding these issues, enabling this technology's long-term operation and commercial viability. However, lowering the operating temperature results in a significant sacrifice in both ionic conductivity and catalytic activity, especially... https://escholarship.org/uc/item/5c90q6xq Mon, 6 Oct 2025 00:00:00 +0000 Barkhordarian, Orbel https://escholarship.org/uc/item/55k892s7 Autonomous out-of-equilibrium agents or cells in suspension are ubiquitous in biology and engineering. Turning chemical energy into mechanical stress, they generate activity in their environment, which may trigger spontaneous large-scale dynamics. Often, these systems are composed of multiple interacting populations that reflect different species, phenotypes, or chemically tailored variants in synthetic settings. Here, we present a new modeling framework for such multi-population active fluids subject to confinement. We adopt a continuum, mean-field, multi-scale description in which each population is characterized by its first three orientational moments. These are dynamically coupled to each other through a shared suspending fluid and to the boundary geometry. The full system is solved using an adaptive, parallel level-set-based method that allows for flexible confinement and efficient resolution of fine-scale structures.Motivated by recent experimental studies, we apply this... https://escholarship.org/uc/item/55k892s7 Mon, 6 Oct 2025 00:00:00 +0000 Fylling, Cayce Julian https://escholarship.org/uc/item/54z4d95j Enduring change is difficult. Many treatments for pathologically canalized experiences and behaviors rely on slow and steady methods focused on structure and narrative. However, it is becoming increasingly clear that slow and steady methods are only weakly effective—the rate of depression, anxiety, PTSD, and other forms of pathological canalization are only increasing, despite our many medical advances. Traditional therapies ameliorate symptoms, but these symptoms invariably return as soon as active treatment ceases. Lasting change requires major neurophysiological and environmental shifts into non-ordinary states. As the mind and world are pulled apart, this opens the door to reconstructing the way you engage with the world. I call this process ‘disintegration-reintegration’. We create critical dynamics to leverage the brain’s ability to learn and relearn.The leading therapies—psychedelic-assisted psychotherapy, focused TMS and ultrasound, EMDR, and emerging solutions like VR-assisted,... https://escholarship.org/uc/item/54z4d95j Mon, 6 Oct 2025 00:00:00 +0000 Clingo, Joshua https://escholarship.org/uc/item/4qs9v99q This dissertation develops and applies a projection method for simulating complex incompressible flows with improved accuracy, simplified data structures, and usability for scientific applications. The method collocates velocity and pressure variables on adaptive octree grids and combines finite volume and finite difference schemes within a level-set framework to resolve complex geometries efficiently. Compared to existing projection methods, it achieves second-order accuracy while streamlining implementation. In the first part of the dissertation, we introduce this method, analyze its stability and convergence, and validate its performance through a series of benchmark tests. In the second part, the method is used to investigate fundamental questions in vortex dynamics, particularly the evolution of vortex lines in transitional flows. In collaboration with experimentalists and theorists at UC Merced, we simulate the destabilization of vortex lines in a pipe flow apparatus, replicating... https://escholarship.org/uc/item/4qs9v99q Mon, 6 Oct 2025 00:00:00 +0000 West, Scott Robert