id |
ndltd-NEU--neu-bz617m20f
|
record_format |
oai_dc
|
spelling |
ndltd-NEU--neu-bz617m20f2021-09-18T05:10:19ZLarge-scale locational data analytics for urban resilienceThere is growing evidence that natural disasters take an increasing toll on the global economy, with a predicted future rise in frequency and intensity under the climate change challenge. In addition to the escalating economic losses, extreme shocks such as pandemics and natural hazards can substantially disrupt human lives, especially in urban areas. Previous studies have attempted to capture, quantify and rationalize such perturbation effects as well as human responses. However, there is still a lack of quantitative understanding of how issues like data quality, pre-existing social and physical disparities, mobility properties, and evolving intensities of extreme shocks can be translated or linked to different emergency behaviors in large-scale emergency contexts. This study aims to advance the understanding of urban resilience with large-scale mobility data and makes three contributions to overcome existing limitations: Firstly, a mobility data quality assurance framework is developed to discover data anomalies issues and three aspects that could contribute to the injustice are identified and examined: representativeness, quality, and precision. Real-world mobility data throughout Hurricane Harvey is used to reveal persistent disparity of representativeness and significant drops of overall data precision throughout the event, which implies that the data biases could be reinforced and perpetuated during extreme shocks, and specific mitigation tasks should be employed. Secondly, I conduct a multi-scale study of disaster-induced evacuations of Hurricane Harvey. By analyzing three-month detailed human mobility data before, during, and after Hurricane Harvey, I report both universality and heterogeneity in multi-dimensional evacuation patterns. The research effort further reveals social disparities in the interplay of risk and resilience. Lastly, under the current challenge of COVID-19, dynamic daily high-resolution mobility networks are constructed and analyzed using methods originated from the network percolation theory. The results demonstrate a large-scale cluster structure with evolving patterns and enable identifying a small, manageable set of recurrent critical links located across the United States, serving as valves connecting divisions and regions. The overall findings provide new insights into understanding the urban social and physical disparities in resilience and managing the connectivity of mobility networks during unprecedented external shocks. --Author's abstracthttp://hdl.handle.net/2047/D20416793
|
collection |
NDLTD
|
sources |
NDLTD
|
description |
There is growing evidence that natural disasters take an increasing toll on the global economy, with a predicted future rise in frequency and intensity under the climate change challenge. In addition to the escalating economic losses, extreme shocks such as pandemics and natural hazards can substantially disrupt human lives, especially in urban areas. Previous studies have attempted to capture, quantify and rationalize such perturbation effects as well as human responses. However, there is still a lack of quantitative understanding of how issues like data quality, pre-existing social and physical disparities, mobility properties, and evolving intensities of extreme shocks can be translated or linked to different emergency behaviors in large-scale emergency contexts. This study aims to advance the understanding of urban resilience with large-scale mobility data and makes three contributions to overcome existing limitations: Firstly, a mobility data quality assurance framework is developed to discover data anomalies issues and three aspects that could contribute to the injustice are identified and examined: representativeness, quality, and precision. Real-world mobility data throughout Hurricane Harvey is used to reveal persistent disparity of representativeness and significant drops of overall data precision throughout the event, which implies that the data biases could be reinforced and perpetuated during extreme shocks, and specific mitigation tasks should be employed. Secondly, I conduct a multi-scale study of disaster-induced evacuations of Hurricane Harvey. By analyzing three-month detailed human mobility data before, during, and after Hurricane Harvey, I report both universality and heterogeneity in multi-dimensional evacuation patterns. The research effort further reveals social disparities in the interplay of risk and resilience. Lastly, under the current challenge of COVID-19, dynamic daily high-resolution mobility networks are constructed and analyzed using methods originated from the network percolation theory. The results demonstrate a large-scale cluster structure with evolving patterns and enable identifying a small, manageable set of recurrent critical links located across the United States, serving as valves connecting divisions and regions. The overall findings provide new insights into understanding the urban social and physical disparities in resilience and managing the connectivity of mobility networks during unprecedented external shocks. --Author's abstract
|
title |
Large-scale locational data analytics for urban resilience
|
spellingShingle |
Large-scale locational data analytics for urban resilience
|
title_short |
Large-scale locational data analytics for urban resilience
|
title_full |
Large-scale locational data analytics for urban resilience
|
title_fullStr |
Large-scale locational data analytics for urban resilience
|
title_full_unstemmed |
Large-scale locational data analytics for urban resilience
|
title_sort |
large-scale locational data analytics for urban resilience
|
publishDate |
|
url |
http://hdl.handle.net/2047/D20416793
|
_version_ |
1719481226296819712
|