Publication Date

2026

Document Type

Dissertation/Thesis

First Advisor

Wilson, James L.

Degree Name

Ph.D. (Doctor of Philosophy)

Legacy Department

Department of Geographic and Atmospheric Sciences

Abstract

The climate is changing, and the livelihoods of more than 3.3 billion people are presumed to be vulnerable to a climate altered significantly from historical norms. Despite international concerns that ramifications of climate change may affect environmental conditions in ways that could exacerbate inundation, dengue fever incidence, and social vulnerability in large areas of the planet, research focused on revealing the roots of these problems tends to cover small areas at large geographic scales. Inconsistencies in research design elements that may ramify from data classification regimen and misappropriation of variable selection, non-consensus about unit of analysis or scale, and disagreement on the choice of a consistent modeling methodology may occur and lead to researcher and information biases that persist in the literature. Therefore, the aim of this dissertation was to develop approaches to modeling inundation susceptibility, dengue fever risk, and social vulnerability models at small geographic scale in Thailand, a country at notable risk to climate change. Each of this dissertation’s projects use ensemble methods, also known as multi-criteria evaluation techniques (MCETs), to identify the most meaningful variables from a raft of options that may potentially have an outstanding effect on phenomena of interest.

This dissertation is divided into three projects. The first project estimates inundation susceptibility for Thailand with the Frequency Ratio Method (FRM), which is a binary statistical modeling approach that is useful for modeling diverse types of environmental risks and disasters. The FRM was deployed to evaluate the importance of 14 independent variables, or “conditioning factors,” on the inundation record of 2005–2020. The resulting inundation susceptibility map was apportioned into discrete maps for each province as a Map Series (ESRI, 2021).

The second project developed a method for generating fine scale raster maps of dengue fever risk based on average monthly incidence rates (per 100,000). Twelve average monthly incidence rates were calculated from monthly province-level case notifications collected during 2003 – 2020 by Thailand’s Bureau of Epidemiology. The average monthly incidence rates were downscaled to constrained urban land use areas with Forest-based Regression (ESRI, 2021). An initial collection of 33 independent variables was evaluated for importance against average monthly dengue fever incidence, of which 19 persisted as meaningful variables that contributed to the incidence of the disease. Variable importance scores derived from each of these 19 variables during the downscaling process were observed, graphed, and applied in developing a map surface of dengue fever risk for the entire country at fine-scale (1 km) with the Raster Calculator (ESRI, 2021).

The third project modeled social vulnerability to climate change in Thailand with the Social-Ecological-Technological Systems (SETS) approach. In this final project, the SETS approach encompasses 18 dimensions of vulnerability and classifies them into three eponymous domains. Each domain was characterized by six distinct variables. As the condition of vulnerability is subjective and uncertain, fuzzy logic was used to delineate geographic areas of vulnerability as a special case of probability with Fuzzy Membership and Fuzzy Overlay (ESRI, 2021). Clusters of vulnerability in each domain were identified with a Moran’s I statistic and the local indicator of spatial autocorrelation (LISA) in the GeoDa software.

Extent

211 pages

Language

en

Publisher

Northern Illinois University

Rights Statement

In Copyright

Rights Statement 2

NIU theses are protected by copyright. They may be viewed from Huskie Commons for any purpose, but reproduction or distribution in any format is prohibited without the written permission of the authors.

Media Type

Text

Download

Share

COinS