Date of Award
Fall 11-16-2024
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Materials and Infrastructure Systems
First Advisor
Elizabeth Matthews
Abstract
Current practice of flood loss prediction presents limitations in accurately predicting building flood losses at multiple scales. While whole-building estimates can more accurately predict high-level losses (i.e., large groups of buildings), a significant analysis error is revealed with small-scale (i.e., individual, or small groups of buildings) investigation. The novel approach of collecting experimental flood data to support damage estimation has provided sufficient evidence that current damage models can and do under or overestimate loss. This research has developed a multi-scale approach to assessing flood losses based on experimentally derived material damage. The methodology is stored and accessible in a multi-scale Flood Loss Assessment Tool (FLoAT). The framework for FLoAT incorporates damage uncertainty and can assess loss costs at multiple scales from individual residential structures to neighborhoods of similar structure types. The model has been refined to only require very few inputs such as flood data (e.g., height of floodwater and type of water), floor area of structure and architectural style depending on the preferred scale of assessment; additional loss costs such as labor, sanitization and restoration costs are also incorporated. The validation of the tool has been completed by analyzing real flood data in Southern Louisiana followed by statistical analysis. The primary goal is accessibility with a strong focus on precision, providing inspectors, city planners, researchers, government officials, and homeowners access to assess damage and loss as precisely as possible.
Recommended Citation
Opel, Anna Katya, "" (2024). Dissertation. 1031.
https://digitalcommons.latech.edu/dissertations/1031
