UTSA Runner Research Press (Univ. of Texas at San Antonio)
Not a member yet
6846 research outputs found
Sort by
Deciphering Eccentric Accretion Disks with GRMHD Simulations
No full textAccretion disks around supermassive black holes play a crucial role in understanding high-energy astrophysical phenomena, particularly in active galactic nuclei (AGNs) and tidal disruption events (TDEs). This study focuses on geometrically thick, eccentric accretion disks and compares them to typically assumed circularized disks used to model advection-dominated accretion flows. Through General Relativistic Magnetohydrodynamic (GRMHD) simulations of Schwarzschild (non-spinning) black holes, we investigate how eccentricity alters accretion dynamics, emission characteristics, and polarization signatures at EHT-relevant frequencies. We analyze weakly magnetized Standard And Normal Evolution (SANE) models, using the KORAL code for fluid dynamics and IPOLE for polarized ray-tracing. Preliminary findings suggest that eccentric disks introduce significant deviations in emission and polarization compared to circularized disks, providing potential observational markers detectable by the Event Horizon Telescope (EHT).Physics and Astronom
Scalable Polar Sea Ice Classification Over Remote Sensing Datasets Using Spatio-Temporal and Machine Learning Techniques
No full textRemote sensing plays a vital role in monitoring and understanding global warming and climate change by providing valuable data about Earth's surface and atmosphere from a distance, typically through satellite-based or airborne sensors.
This research utilizes remote sensing data from ICESat-2 (IS2) and Sentinel-2 (S2) satellites to analyze and forecast key insights into polar regions, focusing on sea ice, freeboard, leads, and open water. By integrating advanced data analysis techniques with GIS computations, the study aims to enhance accuracy and efficiency in handling large-scale spatio-temporal datasets. The goal is to better understand global warming, climate change, and their effects on polar ice, contributing valuable insights into ice extent, thickness, volume, and their role in sea-level rise.
This dissertation presents a series of advancements in polar sea ice classification and freeboard estimation using remote sensing data. First, we introduce a robust system for classifying polar sea ice into categories of thick/snow-covered, young/thin, or open water using Sentinel-2 (S2) imagery from the Ross Sea region in the Summer season. We develop a precise method for segmenting and automatically labeling the S2 images based on carefully defined color thresholds. These auto-labeled datasets are used to train a U-Net machine learning model, resulting in high classification accuracy.
Next, we extend this work by presenting a parallel, distributed method for segmenting and automatically labeling S2 images from the Ross Sea. To scale the auto-labeling process, we implement a parallel workflow using PySpark, substantially improving data loading and map-reduce processing speed. The auto-labeled data are then utilized to train a U-Net model with strong classification accuracy. To address the computational demands of training the U-Net model, we distribute the training across 8 GPUs using the Horovod framework on a DGX cluster, achieving near-linear speedup while maintaining model accuracy.
Then, we generalize the classification approach to be independent of season and region. By modifying our auto-labeling technique for thin cloud and shadow-filtered S2 imagery, we generate high-quality labels for both the Arctic and Antarctic regions during summer and partial winter seasons. This scalable method utilizes a parallel PySpark workflow, and the auto-labeled data are used to train a U-Net model for robust, region- and season-independent sea ice classification. The training process is distributed using Horovod, ensuring high efficiency without compromising classification accuracy.
Furthermore, we shift focus to improving sea surface height and freeboard resolution using ICESat-2 (IS2) ATL03 geolocated photon data. We resample the ATL03 data into 2m segments and classify them into thick ice, thin ice, and open water using deep learning methods, including LSTM and MLP models. Training data is generated through auto-labeling from S2 imagery, with manual corrections for transitions and cloudy regions. After model training, we compute local sea surface heights from open water segments and derive freeboard values.
Finally, a scalable workflow for sea ice classification and freeboard estimation using IS2 ATL03 data. By aligning S2 imagery with IS2 tracks, we create labeled training data for deep learning models, which are trained in a distributed manner using Horovod on a DGX A100 8-GPU cluster. The local sea surface heights are computed from open water segments, and freeboard is scaled with the derived sea level. This scalable workflow leads to significantly improved resolution and accuracy of sea ice classification and freeboard estimation compared to existing ATL07 and ATL10 products.
This research advances our understanding of spatiotemporal dynamics and detects changes in polar regions through detailed analysis of remote sensing datasets. By applying advanced techniques and technology, it provides valuable insights into the factors driving environmental shifts, contributing to the field of remote sensing data analysis and classification.Computer Scienc
Cognitive Computing for Understanding and Restoring Color in Renaissance Art
No full textIn this article, for the first time on this topic, we analyze the historical color palettes of Renaissance oil paintings by using machine-learning methods and digital images. Our work has two main parts: we collect data on their historical color palettes and then use machine learning to predict the original colors of paintings. This model studies color ratios, enhancement levels, symbolic meanings, and historical records. It looks at key colors, measures their relationships, and learns how they have changed. The main contributions of this work are as follows: (i) we develop a model that predicts a painting’s original color palette based on multiple factors, such as the color ratios and symbolic meanings, and (ii) we propose a framework for using cognitive computing tools to recover the original colors of historical artworks. This helps us to rediscover lost emotional and cultural details.Electrical and Computer Engineerin
Sex-Specific Cardiac Effects of Maternal and Offspring High-Fat, High-Sucrose Diets in C57BL/6J Mice
No full textMaternal obesity and exposure to high-fat, high-sucrose (HFHS) diets during development have been linked to long-term cardiometabolic risk in offspring. This study investigated the combined effects of prenatal and postnatal HFHS diet exposure on body composition and cardiac function in male and female mice. Offspring were maintained on the same diets as their dams (control [CON] or HFHS) until tissue collection at 29–32 weeks of age. Body weight and fat mass were significantly increased in HFHS groups, with males showing the greatest adiposity and final body weight. Lean mass was also higher in males, though lean mass relative to body weight declined with HFHS exposure. Cardiac assessments revealed both load-dependent and load-independent impairments. HFHS exposure led to increased end-diastolic and end-systolic volumes, reduced ejection fraction, and lower end-systolic elastance (Ees), consistent with systolic dysfunction. Diastolic function was compromised in a sex-specific manner: male HFHS animals exhibited reduced dP/dt min (relaxation), while female HFHS animals showed increased end-diastolic elastance (Eed), indicating elevated ventricular stiffness. Ventricular-arterial coupling, measured by Ees/Ea ratio, was reduced in HFHS animals of both sexes, with more pronounced impairments in females. These findings highlight sex-specific susceptibilities to HFHS-induced cardiovascular dysfunction, with female offspring exhibiting greater impairments in myocardial contractility and stiffness. This study underscores the need to consider sex as a biological variable in dietary programming research and supports the use of sex-specific approaches in cardiovascular disease prevention.Health and Kinesiolog
Exploring the Voices: A Case Study of the Experiences of Head Start Policy Council Members in a Large Urban Setting
No full textExtensive research emphasizes the benefits of early education programs on child development and school readiness (Henderson & Mapp, 2002; Fantuzzo et al., 2004; Jeynes, 2012; Barnett, 1995; Heckman, 2011; Yoshikawa et al., 2013), yet little attention has been given to how parents and community representatives engage in governance structures like the Head Start Policy Council. As national policies increasingly emphasize parent involvement (Epstein & Sanders, 2006; ESSA, 2015; IDEA, 2004; Texas Education Code § 29.168, 2015), understanding how these governance bodies function and how members experience participatory decision-making is essential.
Head Start, a federally funded initiative, provides comprehensive services to low-income families (Economic Opportunity Act, 1964), yet limited research examines how its Policy Council fosters parental leadership and advocacy (ECLKC, 2023). This study addressed the research gap by exploring how Head Start Policy Council members participate in governance, influence policies, and develop leadership skills using Fung’s (2006) participatory democracy framework.
This qualitative case study employed semi-structured interviews, observations, and document analysis to investigate how members participate in decision-making, develop leadership skills, and overcome barriers such as institutional hierarchies and limited training opportunities. Findings revealed that Policy Council members gained confidence, improved their communication, and developed advocacy skills, which positively influenced program policies. However, structural challenges often hindered engagement. Despite these barriers, participants reported increased empowerment and collaboration. By amplifying the voices of Policy Council members, this study contributes to the discourse on family engagement, shared governance, and participatory democracy in early childhood education.Educational Leadership and Policy Studie
Implications and Integration of Religion/Spirituality in CACREP Counselor Education Programs for Counselors-In-Training
No full textThis study examines the integration of religion and spirituality (R/S) within counselor education programs accredited by the Council for Accreditation of Counseling and Related Educational Programs (CACREP). Specifically, this research explores counselors-in-training (CITs) and their competency, preparedness, and confidence in addressing R/S issues in counseling practice. Data collection and analysis focusing on CITs' exposure to R/S topics during graduate coursework, their perceived significance of the Association for Spiritual, Ethical, and Religious Values in Counseling (ASERVIC) Spiritual Competencies, and their self-assessed ability to implement these competencies effectively. Additionally, the study investigates CITs' comfort levels in applying these competencies across diverse client populations, as well as their perspectives on potential harm and trauma resulting from R/S values conflicts in counseling settings. The findings contribute to the discourse on counselor education by highlighting the necessity of comprehensive R/S training to enhance ethical and culturally responsive counseling practices.Counselin
Reactivity of Z-3-Hexenal with Amino Groups Provides a Potential Mechanism for Its Direct Effects on Insect Herbivores
No full textGreen leaf volatiles (GLV) have been shown in the past to significantly affect herbivore performance in plants by activating direct and indirect defense measures. At the same time, insect herbivores have developed multiple mechanisms to reduce the quantities of Z-3-hexenal, the first product of the pathway. Among the various measures is the abundance of a molecule named hexenal trapping molecule (HALT), which appears to bind large quantities of freshly produced Z-3-hexenal (Z3al). Functional groups like -NH<sub>2</sub> within HALT may be responsible for the binding of Z3al. To test for this potential interaction, we tested different amino acids in various binding assays for their capacity to bind Z3al. We found a significant reduction in Z3al production in the presence of these amino acids, presumably through Schiff base formation. In a feeding assay with beet armyworm (<i>Spodoptera exigua</i>), we found a significant impact of a Z3al- or E2al-spiked artificial diet on growth and development. Together, these data indicate that the presence of Z3al in a diet can harm insect herbivores, which may help to further explain the efforts insect herbivores undertake to suppress the biosynthesis of this compound.Biology, Health, and the Environmen
Benchmarking Container Technologies for ARM-Based Edge Computing
No full textIoT devices continuously send data to the cloud for processing. However, with billions of devices expected, this impacts performance and cost. Processing data locally in edge nodes closer to IoT devices can improve system performance by computing tasks in edge devices and reducing cloud data transfer. However, Edge devices are generally resource-constrained with limited RAM, CPU, and Storage; thus, the efficient use of edge devices is necessary. Container technologies are ideal for edge computing due to their isolation, lightweight nature, and portability. However, without a standardized framework, comparing different container technologies, architectures, and applications is challenging. While benchmarking schemes for containers on edge devices can address these issues, there is limited research in this direction. So, in this work, we take a step towards developing benchmarking schemes that will lay the foundations for the widespread use of benchmarks in edge computing, guiding edge system designers and users in developing the best edge system for their needs.
We experiment with containers using existing synthetic benchmark tools, including Sysbench, Apache Benchmark, Iperf, and Stress-ng. Despite their usability in benchmarking a single system, they do not portray the effectiveness of multiple containers managed in an edge environment. Accordingly, we conduct real-world benchmarking experiments evaluating computer vision applications and data science workloads commonly encountered in edge computing scenarios. Based on the results, we define a set of metrics for benchmarking containers in edge computing. In the proposed benchmarking schemes, we explore and evaluate the performance, efficiency, and suitability of different container technologies, including Docker, Podman, Singularity, and LXC, in the context of edge computing on ARM-based devices. Based on our findings, we offer practical advice for picking the best container technology for specific use cases in ARM-based edge computing.
Besides, we extend the benchmark tool to work with benchmarking container orchestration tools, such as Docker Swarm, Kubernetes, and Nomad. Since edge systems involve multiple edge nodes, orchestration improves the efficiency of edge systems. Moreover, Unikernel has a faster boot time, a smaller footprint, and flexibility, which can be an excellent potential for edge computing. We study their feasibility in Edge computing. Furthermore, we implemented a hybrid edge system using containers and unikernels based on the application type and resource efficiency.Computer Scienc
Urban Open Space Resilience and Social Vulnerability to Seismic Risks: A GIS-Based Analysis of San Francisco
No full textThe thesis aims to explore urban open space resilience through an examination of the relationship between physical and social vulnerability of San Francisco census tracts with regards to earthquake risk. Social vulnerability data was obtained from the CDC Social Vulnerability Index. An open space resilience score was calculated using a combination of measurements that were obtained using open data from the City and County of San Francisco and OpenStreetMap data. Two phases of analysis were performed. Each phase consisted of ANOVA tests conducted to assess the relationship between social vulnerability and various elements of physical vulnerability. The ANOVA tests revealed that significant differences exist between the average open space resilience scores across social groups in San Francisco, specifically with regards to Housing Type & Transportation and Racial & Ethnic Minority Status. Additionally, physical factors like tall building distance, open space size, and service center distance differed significantly between social groups. This research clarifies the ways in which the built environment of the city diminishes or strengthens open space resilience to earthquake events. Planning implications include the construction of vulnerable housing away from hazard zones, the inclusion of earthquake resilience as a guiding principle of open space development, and greater communication clarity of the benefits and uses of open spaces in earthquake response and recovery across levels of governance. This research contextualizes earthquake resilience in the greater socioeconomic landscape of the city, illuminating opportunities for equity improvements and community engagement.Urban and Regional Plannin
Aztlán en Movimiento: An A/r/tographic Inquiry on M.E.Ch.A. as Sponsor of Identity Development for Student Activists
No full textThe full text of this item is not available at this time because the author has placed this item under an embargo until May 15, 2027.Ethnic-based spaces in higher education offer the type of vital academic, emotional, and culturally affirming support that many colleges and universities often overlook for Latinx students. Emerging from the Chicano Movement, M.E.Ch.A. is one of the most iconic and enduring Latinx student organizations. However, little is known about how M.E.Ch.A., the organization, serves as a sponsor of thirdspaces for MEChistAs and its influence on their ethnic identity development. Sponsorship is a relatively new concept in higher education. It has recently gained attention for its positive outcomes for minoritized students in doctoral and leadership programs. It is essential to distinguish that while mentorship guides personal relationships, sponsorship invests strategically to advance the protégé’s interests. Using a/r/tography, an innovative arts-based method, I combined collage-making with historical archival and graphic content from M.E.Ch.A.’s past newsletters and other significant publications from the Chicano Movement to examine the experiences of MEChistAs. Findings indicate that M.E.Ch.A. serves as a sponsor of thirdspaces for connection with the community, cultural pride, and collective resistance. Findings also suggest that M.E.Ch.A. operates as a sponsor of identity development for MEChistAs by shaping their identities and positively transforming how they understand themselves. To close, I presented recommendations for faculty, staff, and administrators, recommendations for future research, and policy implications.Educational Leadership and Policy Studie