Semester of Graduation

2019 Spring

Degree

Master of Science (MS)

Department

Department of Geography and Anthropology

Document Type

Thesis

Abstract

Resilience is a concept with increasing importance in modern risk management because of its role in reducing risks of unpreventable disasters. Previous resilience assessment studies often require extensive surveys of various social, economic, and psychological data or incorporate remote sensing data as one of the complicated physical and social parameters for assessment models. Limited data accessibility to such data due to funding, time, and labor intensity is a major challenge for their wider applications. Therefore, this study proposes the hypothesis that the overall resilience of an urban area to disturbances of natural disasters can be reflected through the time series change sequences of thermal and vegetation index from satellite images. This is because the vegetation index reflects the recoverability of vegetated areas, and thermal change pattern is a reflection of land-cover and land-use changes and energy consumption, which is the end result of various impacts such as social, economic, and physical factors.

Specifically, this study introduced a rapid and objective flood resilience assessment method through time series classification based on thermal feature and vegetation index. The method first used unsupervised classification methods to identify potential flood impact levels and conducted supervised classification to obtain a more accurate classification result. Finally, the derived impact levels were classified as flood resilience levels, which are beneficiary for flood preparation and resource allocation for the local and federal government.

Committee Chair

Meng, Xuelian

DOI

10.31390/gradschool_theses.4866

Download

Share

COinS