UW-Madison Department of Geography Master's Theses

Fine-Scale Vegetation Change and Its Implications for Methane Emissions in Arctic Aquatic Ecosystems

2025-10-23T11:38:58Z

Fine-Scale Vegetation Change and Its Implications for Methane Emissions in Arctic Aquatic Ecosystems Maraldo, Daniel J. While small water bodies (ponds) make up a small fraction of the surface area of Arctic permafrost landscapes, they have been observed to emit disproportionately high volumes of methane and carbon dioxide. To assess drivers of emissions, we mapped the spatial distribution of two aquatic graminoid species associated with high carbon fluxes in Utqiagvik, Alaska over a two-decade period. We combined in situ carbon flux observations with very high resolution panchromatic satellite imagery and drone photogrammetry to directly assess changes in 255 ponds across 5 drained thaw lake basins. Across the study, aquatic graminoid species significantly diminished in spatial coverage. We found that larger ponds (<330m2) drove the trends in vegetation cover loss, while smaller ponds remained stable. These changes also occurred mostly in the past decade (2012-2023). The rapid loss of aquatic emergent vegetation cover could stem from shifts to hotter and dryer climate, increases in pond temperature, and active layer depth. Given the role of methane transport by aquatic plants, our findings represent a significant change in methane emissions and a potential shift in the pathways of carbon in aquatic ecosystems in the Barrow Peninsula. A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science (Geography) at the University of Wisconsin-Madison.

Driftless Divided: Cardinal-Hickory Creek and Wisconsin Transmission Resistance

2025-10-23T11:38:57Z

Driftless Divided: Cardinal-Hickory Creek and Wisconsin Transmission Resistance Shapiro, Gabriel Noah Introduction to U.S. Transmission and Transmission Resistance: The U.S. is experiencing a renewable energy building boom. Transmission lines, along with wind and solar farms, are being proposed and built, across the country. In order to unlock renewable potential, most abundant in the nation’s center, we have been told, we need to build high voltage transmission lines (HVTLs) to get electricity to where it’s needed. These are not small projects and their price tags are growing. As public serving infrastructure, how they’re built matters. The way they’re designed will contribute to how they engage with, and serve society over time. Circulating discourses posit that “we need more HVTLs,” but often stop there, not asking, “what kind of HVTLs do we need?” Rural communities along the path of these projects often resist them, and as a result, developers and some supporters of renewable energy blame those communities, not only for slowing down individual projects but for slowing down the national transition to renewable energy. They are often depicted as either self-centered NIMBY’s (Not In My BackYard), overly pure “tradeoff denying” environmentalists, or “angry farmers” who just don’t like change, don’t know what they’re talking about, and aren’t making productive suggestions for alternatives. A narrative of “green civil war” depicts farmers and conservationists at war with renewables. This transition does need to happen quickly, in order for us to effectively address climate change. However, this common set of assumptions might actually be slowing the transition down more. Expecting the “barrier” of those kinds of opposition, HVTL developers often come into communities defensively, and at the last minute, delivering information selectively, and in one direction, an approach which ends up perpetuating conflict, lawsuits, and delays. A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science (Geography / Environment & Resources) at the University of Wisconsin-Madison, 2025.

Mitigation of Spatial Bias in Social Media Data for Disaster Relief

2025-10-22T10:55:26Z

Mitigation of Spatial Bias in Social Media Data for Disaster Relief Dhakal, Ashmita Social media platforms generate large volumes of geotagged data that are becoming more valuable for disaster response and situational awareness. However, the opportunistic and skewed nature of Volunteered Geographic Information (VGI) makes it spatially biased, making such data less representative and distorting disaster relief efforts. This study proposes a new method to mitigate the effects of this spatial bias in geotagged social media data, specifically tweets made during Hurricane Sandy in New York City. The study uses environmental covariates like building density, elevation, distance to shoreline, and Modified Normalized Difference Water Index (MNDWI) to measure the representativeness of tweet samples in each covariate space. Two optimization techniques, Genetic Algorithm (GA) and Stochastic Gradient Descent (SGD)-based optimization, were used to determine an optimal weight for individual tweets to adjust their spatial representation by improving their distribution in the covariate space. This study found that the mitigation process is effective and that the SGD-based approach significantly outperformed GA, with overall similarity scores increasing from 0.679 (initial) to 0.895. The improvement was particularly notable in flood-sensitive areas, as evidenced by the MNDWI similarity score rising from 0.605 to 0.948. The weighted tweet maps generated using the optimized weights identified the disaster-affected areas more effectively and were in close agreement with the official inundation data published by FEMA. This research contributes a scalable and data-driven framework for spatial bias mitigation in social media data and helps enhance the reliability of such data in post-disaster mapping and response planning. A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Cartography & Geographic Information Systems at the University of Wisconsin-Madison, 2025.

Mapping Urban Coyote Ecology in Los Angeles: Insights from Citizen Science and Human Mobility Data

2025-10-22T10:55:25Z

Mapping Urban Coyote Ecology in Los Angeles: Insights from Citizen Science and Human Mobility Data Zhang, Qianheng Understanding how urban coyotes (Canis latrans) respond to human activities is a critical challenge in urban ecology, especially in an era of rapid urbanization. As coyotes adapt to urban environments, the list of citizen reports on coyote occurrence together with their locations also becomes more frequent and diverse, offering new opportunities to study their behaviors on a larger scale for human-coyote interaction. This study investigates the spatial and temporal distributions of coyotes in Los Angeles County by integrating citizen science data from iNaturalist together with environmental, socioeconomic and human mobility datasets. We develop a species distribution model using Random Forest and Geographically Weighted Regression(GWR) to identify key ecological and anthropogenic drivers. Furthermore, we employ structural equation modeling (SEM) to explore how time-varying human visitor flows, particularly during the Covid-19 pandemic, influence urban coyote visibility across neighborhoods. Our findings reveal that spatial patterns of coyote occurrence are strongly influenced by environmental and socioeconomic variables. The Random Forest and GWR models highlight that socioeconomic conditions such as poverty rate and population density are key predictors of the use of coyote habitat, with lower income and high density areas showing higher incidence. Furthermore, the spatial heterogeneity in the correlation between seasonal environmental factors and socioeconomic variables reflects the adaptive habitat selection strategies of coyotes at different times of the year. SEM further reveals that coyote observations increase significantly with human inflow in real time during and after the pandemic, while declining in response to sustained human absence. This suggests that coyote behavior is more shaped by short-term human mobility patterns than by long-term redistribution. Importantly, we demonstrate that citizen science data, while subject to reporting biases, correlate strongly with ecological suitability and human mobility patterns, offering a unique perspective on urban wildlife dynamics using spatial data science approaches. A dissertation submitted in partial fulfillment of the requirements for the degree of Master of Science (Geography) at the University of Wisconsin-Madison, 2025.