18431 research outputs found
Sort by
The Ballot
Get PDFThe Ballot is a political fantasy set in the nation of New Salisar, where five states— divided by history, violence, and ideology—teeter on the edge of collapse. In a rare act of unity, the states agree to a radical solution to choose their next leader, a high-stakes competition called the Ballot. Run by the leader of The Vultures, Lucius, the Ballot starts as a symbol of hope—until it turns deadly.
At the center of the story are two contenders: Leon, the original heir to the thron and Serafina, a bold girl from the so-called “rebel” state. The relationships between the six contenders in the Ballot are filled with secrets, alliances, and betrayal. As candidates clash in a series of challenges meant to test their ability to lead New Salisar, a darker plan unfolds beneath the surface. This story explores how real change in government often needs more than action— it needs a powerful story to carry it
Invisible No More: Looking and Listening Locally to Build Conflict-Related Sexual Violence Resilience
Get PDFThe ubiquity of sexual violence, with its devastating physical and psychological consequences, has rendered conflict-related sexual violence (CRSV) one of the most effective and enduring “invisible crimes” of war. Despite receiving growing attention across academic, political, and activist circles, wartime sexual violence remains pervasive in contemporary conflict zones. Against the backdrop of these frequent and recurring violations, this thesis explores the dissonance between high-level rhetorical commitments to preventing and punishing CRSV and on-the-ground realities. What soon emerges is a troubling pattern in which gendered violence—and women’s lived experiences more generally—while increasingly visible in theory, remains marginalised in practice. By investigating the limitations of existing frameworks, this thesis explores why current efforts have failed to mitigate sexual and gender-based crimes in conflict settings. Through a series of semi-structured interviews with humanitarian workers, this study introduces the concept of “preparedness” as a critical factor in strengthening CRSV resilience, emphasising the importance of context-specific, gender-sensitive, and survivor-centred interventions across all stages of the conflict cycle. It concludes by outlining several policy recommendations for future practice and theory-building related to CRSV preparedness
Collision Avoidance and Vegetation as Drivers of Collective Motion in Australian Plague Locusts
Get PDFThe Australian plague locust (Chortoicetes terminifera) is an agricultural and ecological pest that causes tens of millions of dollars in crop damage each year. In this thesis, we develop mathematical models of hopper bands, dense formations of juvenile locusts that move across a vegetated field and destroy plants in their path. We develop agent-based and PDE models of hopper bands moving through vegetation to examine how recently discovered behavior where locusts slow down to avoid collisions with other locusts may influence the shape, speed, and destructiveness of hopper bands. We find that collision avoidance may cause locusts to aggregate at higher densities in some parameter regimes, making collision avoidance a possible contributing factor to why locusts aggregate at densities as high as 4000 per square meter. Our results may advance scientific understanding of collective structures in locusts and could benefit locust control efforts, which are essential to maintaining food security and economic productivity and will become more crucial in a warming world
Batteries for Grid Stability: The Impact of Utility-Scale Battery Installation on Electricity Price Volatility in California
Get PDFAs renewable energy penetration increases, driven largely by solar, electricity prices have become increasingly volatile due to the intermittency of supply, increasing the need for price stabilizing interventions. This study examines the impact of cumulative additions of utility-scale battery energy storage systems (BESS) capacity on real-time market (RTM) electricity price volatility within the California Independent System Operator (CAISO) between 2020 and 2024, taking a long-term, system-wide perspective on storage deployment. Using an Ordinary Least Squares (OLS) regression framework, the analysis finds that cumulative BESS capacity is significantly associated with reduced price volatility in the benchmark model, with results remaining robust across multiple specifications.
Notably, the study finds that the interaction between cumulative BESS and natural gas prices is consistently positive and significant, suggesting that storage is more effective at mitigating volatility when gas prices are low. These results imply that BESS may not be effectively reducing volatility stemming from external factors such as gas price shocks and may primarily target volatility caused by renewable intermittency. As a result, BESS may not be fully capitalizing on available price arbitrage opportunities on the grid, highlighting the importance of considering fuel price dynamics when evaluating BESS performance
The Effect of the TCJA and CARES Act on Charitable Contributions and the Financial Health of the Nonprofit Sector
Get PDFThis thesis examines the impact of the Tax Cuts and Jobs Act of 2017 and the Coronavirus Aid, Relief, and Economic Security Act of 2020 on charitable contributions and nonprofit financial health. The standard deduction increase introduced by the TCJA disincentivized middle to low-income earners from making contributions. In March 2020, Congress temporarily increased the AGI limit on charitable deductions while simultaneously offering an above-the-line deduction for all U.S. taxpayers. While earlier studies have analyzed the effect of the TCJA in disincentivizing charitable contributions, no study has examined the moderating effect of the TCJA and CARES Act in affecting contributions to smaller nonprofits. As middle-to-low-income earners tend to donate to and benefit from smaller organizations, this thesis examines the effect of both the TCJA and CARES Acts in incentivizing charitable contributions and their preceding effects. The analysis focuses on the effects of smaller nonprofits eligible to file a 990EZ form, and the outcomes of this legislation by examining contributions, employee counts, volunteer activities, net assets, and cash balances. Utilizing a panel dataset from IRS Form 990 Filings between 2014 and 2021, I conduct a difference-in-differences framework to analyze the effects of these policies. The results show a significant reduction in individual contributions to smaller nonprofits both post-TCJA and CARES compared to their larger counterparts. These findings highlight the increase in standard deduction, while reducing tax liability, reduced charitable contributions, and the financial well-being of the nonprofit sector, as well as show the sensitivity of nonprofit funding to federal policy
The Impact of FEMA-Declared Disasters on U.S. Home Values
No full textThis thesis examines the impact of FEMA-declared disasters on home values in the United States, using the annual change in the FHFA House Price Index (HPI) as the primary outcome. I find that a FEMA-declared disaster leads to a decrease of 0.642 in the annual change in HPI two years after the event, reflecting delayed housing market frictions. The analysis then examines FEMA disaster aid’s role in post-disaster outcomes. A full increase in the ratio of aid received to damage assessed (from no coverage to full coverage) is associated with a 0.779 decrease in the annual change in HPI during the disaster year, likely reflecting underlying damage severity rather than aid effects. Conversely, a full increase in the ratio of individuals receiving aid to those with valid registrations is associated with a 1.61 increase in the annual change in HPI four years after the disaster, suggesting broader aid access supports long-term market stabilization. The study also evaluates two major policy reforms: the Sandy Recovery Improvement Act (SRIA) of 2013 and the Disaster Recovery Reform Act (DRRA) of 2018. I find that ZIP codes experiencing disasters after SRIA saw an increase of 0.731 in the annual change in HPI during the disaster year, while the DRRA had no significant effect. Finally, I examine political influences on aid distribution. Although Democratic- voting counties had a 1.75% lower approval rate for FEMA registrations, there was no significant difference in aid awarded, suggesting disparities stem from registration patterns rather than systematic bias
The Role of Maternal Haploid in Nasonia Vitripennis Development
No full textKey chromatin remodeling events occur during spermatogenesis and during fertilization and are essential for proper incorporation of the paternal genome. In Drosophila melanogaster, the maternal effect mutation maternal haploid (mh) disrupts this process, leading to paternal genome elimination (PGE) and the development of gynogenetic embryos. In this study, we investigated the role of the mh ortholog in the parasitoid wasp Nasonia vitripennis. We performed RNA interference (RNAi) knockdowns of MH in both females and males. We showed that crosses between mh females and wildtype (WT) males yielded normal sex ratios and embryos with normal nuclei, indicating a functional divergence or an insufficient knockdown. In contrast, sex ratios observed in male knockdown and WT female crosses were half normal and half all-male broods. These results indicate that MH is required for producing functional sperm or that these broods were a product of virgin females. We were also interested in testing whether MH is required for the activity of PSR, a selfish B chromosome that induces PGE, but were unable to carry out these experiments. Future work will include more extensive knockdown validation, immunostaining to determine MH localization in embryos and germlines, and functional experiments involving PSR. Together, these studies aim to clarify the role of MH in N. vitripennis and deepen our understanding of chromatin remodeling and genome elimination
Climate Change and Ski Resort Ownership Consolidation
Get PDFBased on a collected dataset of ski resort ownership records and localized climate data spanning 1920-2020, I analyze how environmental volatility affects firm-level outcomes. Using multinomial logit regression models, I examine whether changes in snowfall, precipitation, and temperature are associated with higher probabilities of acquisition, particularly by large conglomerates. Preliminary results suggest that climate volatility, specifically around minimum and maximum temperature, is becoming increasingly important in affecting ownership transfers. Resorts with higher volatility in warm-day temperatures are less likely to be acquired by both conglomerates and non-conglomerates, while cold volatility has become less important over time as snowmaking capabilities are increasing. Overall, the findings indicate that climate volatility, rather than average climate levels, are more critical to influencing ownership transfers, as heterogeneity across the ski resort industry results in slightly different adaptation strategies
The Coral Porites lutea as a Regional and Local Proxy for Sea Surface Temperature in Hong Kong, and the Effects of Typhoons on Proxy Accuracy
Get PDFProxies related to the climate have been used as a method to observe long-term changes occurring underwater that span long before we had the capability to measure environmental characteristics. Porites corals have been used as sea surface temperature (SST) proxies due to annual cyclicity in their growth process and mineral composition. As the climate differs globally, it is important to ensure that proxies are accurate regardless of location to measure changes in climate. This study examines Porites lutea corals as an SST proxy in the region of Hong Kong. Potential local differences in SST proxy calibrations within the surrounding waters of Hong Kong are examined, as the inlets, tropical weather, and large human population create a highly active and volatile environment. Due to the intense wet seasons experienced in Hong Kong, typhoons are included as a covariate and interaction term in the proxy calibrations to test for significant influence in the correlational relationship. Age estimates for each coral sample are determined using a combination of mapped chronologies and data analysis. All age estimates are cross-referenced with estimated growth rates in past Porites Hong Kong studies to ensure accuracy. The SST proxy calibrations are estimated using linear, ANCOVA, and mixed effects models. The ratio of Strontium isotope 88 and Calcium isotope 43 (Sr/Ca) is a reliable SST proxy for the region of Hong Kong, at an overall slope of -0.606 mmol/mol per °C (p \u3c 2.82e-06). Between the three sites sampled in this study, two have statistically significant different proxy slopes. As these sites are the farthest apart, this indicates that regions may need more localized SST proxy calibrations, especially if there are highly variable and locally isolated environments within the region. Additionally, all sites exhibit different intercepts, indicating that local conditions influence the baseline Strontium and Calcium quantities in corals. Typhoons do not significantly impact Sr/Ca-SST proxy calibrations in Hong Kong, but the explained variation in the correlational model is still relatively low (R2 = 0.606). This indicates that variables unrelated to weather--such as human impact, or proximity to land--should be included instead in future studies
DQN Reinforcement Learning Approaches for the Combinatorially Complex game of Checkers
No full textBoard games have long served as foundational testbeds in Reinforcement Learning (RL) research, offering structured environments to train, test, and benchmark agents. By abstracting key elements of real-world decision-making, such as strategic planning, resource management, uncertainty, and competition, these games provide a simplified yet meaningful platform for experimentation. As a result, board games have become a gold standard for facilitating fair algorithmic comparisons in RL. This paper investigates several approaches for training a Deep Q-Network (DQN) agent to learn the game of Checkers, examining four distinct learning setups: training against a random agent, against a Minimax agent, against a curriculum-based ensemble of opponents, and through pure self-play. The study details the design decisions involved in modeling the Markov Decision Process (MDP) environment, constructing opponent strategies, structuring agent training, and ultimately analyzes the experimental results to assess the strengths and limitations of each method