Author ORCID Identifier:
Date of Graduation
7-2026
Document Type
Dissertation
Degree Name
Doctor of Philosophy in Environmental Dynamics (PhD)
Degree Level
Graduate
Department
Environmental Dynamics
Advisor/Mentor
Aly, Mohamed
Committee Member
Matlock, Marty
Second Committee Member
Lewis, Sarah
Keywords
Flood damage; Flood risk assessment; GeoAI
Abstract
Flooding is one of the most recurrent and economically destructive natural hazards globally, with annual losses estimated at approximately $50 billion and accelerating under the combined pressure of climate change and rapid urbanization. In Arkansas, the convergence of major river systems, rapidly growing metropolitan areas, and socioeconomically vulnerable communities creates an urgent and underserved need for advanced, locally grounded flood risk science. This dissertation addresses that need through an integrated, three-component geospatial research program encompassing systematic evidence synthesis, scenario-based hydraulic loss modeling, and GeoAI-driven event-scale impact assessment. The first component delivers a comprehensive systematic review of contemporary geospatial approaches to flood risk and hazard assessment, synthesizing 89 peer-reviewed articles published over the past 15 years using the PRISMA framework. The review maps the methodological landscape across three primary domains, GIS and remote sensing, hydrological and hydraulic modeling, and machine learning, and documents persistent gaps including the chronic underrepresentation of rural flood risk and incomplete integration of hazard, exposure, and vulnerability into unified risk frameworks. These findings establish the conceptual and methodological scaffolding for the empirical contributions that follow. The second component presents integrated flood scenario modeling and loss estimation for Northwest Arkansas (NWA) using the HAZUS-MH platform. Flood depth grids for 10-, 20- , and 100-year return period events were developed for the four-county NWA region and used to estimate Annualized Average Loss (AAL) at the census-tract level. Total regional AAL is estimated at $13.89 million annually, with residential structures accounting for 47.6% of losses. Validation against FEMA's National Risk Index (NRI) reveals that the NRI systematically underestimates flood risk by 32.5% at the regional scale, with the most severe discrepancy in Carroll County where a gap of $3.39 million — representing 82.4% of actual AAL — is attributable to aging building stock, complex White River–Beaver Lake hydraulics, and concentrated low-income populations with less than 8% flood insurance enrollment. The third component presents a GeoAI framework for flood mapping and multi- dimensional impact assessment, applied to the 2019 Arkansas River flood event in Conway County. Sentinel-1 Synthetic Aperture Radar (SAR) imagery was processed in Google Earth Engine using a pretrained deep learning model to produce a flood inundation map at 12.67-meter resolution, achieving 91.7% overall accuracy and a Kappa coefficient of 0.83. HAND terrain- based validation confirmed 99.7% agreement between observed inundation and predicted risk zones. Multi-dimensional impact analysis revealed that 80.77 km² were inundated, with agricultural land (cropland and pasture) accounting for 62.7% of total flood area. NDVI anomaly detection identified 5.26 km² of severe crop damage, and post-flood urban expansion analysis documented a 16.4% increase in developed area between 2019 and 2021, with high-intensity development growing by 195%, signaling a compounding future flood risk trajectory. Collectively, these three studies demonstrate the value of integrated, multi- methodological geospatial frameworks for flood risk science in rapidly changing landscapes. The findings provide actionable evidence for flood risk governance in Arkansas, with methodologies that are reproducible, scalable, and directly transferable to other flood-prone regions globally.
Citation
Gasmi, F. (2026). An Integrated Geospatial Framework for Flood Risk Assessment, Loss Estimation, and Impact Analysis in Arkansas. Graduate Theses and Dissertations Retrieved from https://scholarworks.uark.edu/etd/6394
