Tracking Disaster Dynamics for Urban Resilience: Human-Mobility and Semantic Perspectives

Fostering urban resilience and creating agility to disaster response is an urgent task faced by cities worldwide in the context of climate change and increasing frequencies of natural disasters. Understanding and tracking the dynamic process of resilience to disasters is the first step to operationalize the concept of urban resilience. In this dissertation, I present four related but evolutionary perspectives to investigate the impact of natural disasters on interactive human-environment systems as well as the dynamic process of resilience, including human mobility, spatial networks, and coupled mobility and sentiment perspectives. In the first, human-mobility perspective, I examine the nuanced impact of a severe winter storm on human mobility patterns and the relationship between perturbed mobility during the storm and recurrent mobility under normal circumstances. In the second, where I adopt a spatial network perspective, I investigate the dynamic process of resilience over time by analyzing networked human-spatial systems using an ecology-inspired approach. The third perspective involves sentiment as an additional factor to human mobility to understand urban dynamics during an earthquake. In this perspective, I explore the relation between disaster magnitude and a population's collective sentiment, as well as temporal correlations between sentiment and mobility. Each of the three empirical studies employs a quantitative, empirical research methodology and uses voluntarily reported geo-referenced data collected through a Twitter Streaming API. After multiple investigations on diverse types of natural disaster (e.g. severe winter storm, flooding, hurricane, and earthquake), I develop a Detecting Urban Emergencies Technique (DUET), as the fourth part of my dissertation, for identifying and tracking general types of emergencies in a short period without prior definitions of emergent topics. Research findings from the three empirical studies and the proposed DUET detection technique introduce a new lens and approach for understanding population dynamics and achieving urban resilience. This dissertation contributes to a more complete understanding of urban resilience to disasters with crowdsourced data, and enables more effective urban informatics in the face of extreme events.

Name thesis:degree_name

PHD

Level thesis:degree_level

doctoral

Discipline thesis:degree_discipline

Civil Engineering

Department dc:contributor.department

Civil and Environmental Engineering

Grantor dc:publisher

Virginia Tech

Year dc:date.issued

2018

Author dc:creator

• Wang, Yan

Chairs dc:contributor.committeechair

• Taylor, John E.
• Garvin, Michael J.

Committee members dc:contributor.committeemember

• Wang, Qi
• Moore, Ignacio T.
• Shealy, Earl W.

• Big Data
• Disaster Dynamics
• Emergency Detection
• Human Mobility
• Sentiment
• Twitter
• Urban Resilience

Statement dc:rights

• In Copyright

Licence dc:rights.uri

• http://rightsstatements.org/vocab/InC/1.0/

Dc Identifier Other

vt_gsexam:15462

OAI identifier oai:identifier

oai:vtechworks.lib.vt.edu:10919/95913