Hyper-resolution monitoring of urban flooding with social media and crowdsourcing data

Ruo Qian Wang, Huina Mao, Yuan Wang, Chris Rae, Wesley Shaw

Research output: Contribution to journalArticle

30 Citations (Scopus)
239 Downloads (Pure)

Abstract

Hyper-resolution datasets for urban flooding are rare. This problem prevents detailed flooding risk analysis, urban flooding control, and the validation of hyper-resolution numerical models. We employed social media and crowdsourcing data to address this issue. Natural Language Processing and Computer Vision techniques are applied to the data collected from Twitter and MyCoast (a crowdsourcing app). We found these big data based flood monitoring approaches can complement the existing means of flood data collection. The extracted information is validated against precipitation data and road closure reports to examine the data quality. The two data collection approaches are compared and the two data mining methods are discussed. A series of suggestions is given to improve the data collection strategy.

Original languageEnglish
Pages (from-to)139-147
Number of pages9
JournalComputers and Geosciences
Volume111
Early online date8 Nov 2017
DOIs
Publication statusPublished - Feb 2018

Fingerprint

flooding
Monitoring
Risk analysis
monitoring
Application programs
Computer vision
Data mining
Numerical models
Processing
computer vision
data mining
social media
data quality
road
Big data

Cite this

@article{967fea975be2404e9ff34a49e84c7fc5,
title = "Hyper-resolution monitoring of urban flooding with social media and crowdsourcing data",
abstract = "Hyper-resolution datasets for urban flooding are rare. This problem prevents detailed flooding risk analysis, urban flooding control, and the validation of hyper-resolution numerical models. We employed social media and crowdsourcing data to address this issue. Natural Language Processing and Computer Vision techniques are applied to the data collected from Twitter and MyCoast (a crowdsourcing app). We found these big data based flood monitoring approaches can complement the existing means of flood data collection. The extracted information is validated against precipitation data and road closure reports to examine the data quality. The two data collection approaches are compared and the two data mining methods are discussed. A series of suggestions is given to improve the data collection strategy.",
author = "Wang, {Ruo Qian} and Huina Mao and Yuan Wang and Chris Rae and Wesley Shaw",
year = "2018",
month = "2",
doi = "10.1016/j.cageo.2017.11.008",
language = "English",
volume = "111",
pages = "139--147",
journal = "Computers and Geosciences",
issn = "0098-3004",
publisher = "Elsevier",

}

Hyper-resolution monitoring of urban flooding with social media and crowdsourcing data. / Wang, Ruo Qian; Mao, Huina; Wang, Yuan; Rae, Chris; Shaw, Wesley.

In: Computers and Geosciences, Vol. 111, 02.2018, p. 139-147.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Hyper-resolution monitoring of urban flooding with social media and crowdsourcing data

AU - Wang, Ruo Qian

AU - Mao, Huina

AU - Wang, Yuan

AU - Rae, Chris

AU - Shaw, Wesley

PY - 2018/2

Y1 - 2018/2

N2 - Hyper-resolution datasets for urban flooding are rare. This problem prevents detailed flooding risk analysis, urban flooding control, and the validation of hyper-resolution numerical models. We employed social media and crowdsourcing data to address this issue. Natural Language Processing and Computer Vision techniques are applied to the data collected from Twitter and MyCoast (a crowdsourcing app). We found these big data based flood monitoring approaches can complement the existing means of flood data collection. The extracted information is validated against precipitation data and road closure reports to examine the data quality. The two data collection approaches are compared and the two data mining methods are discussed. A series of suggestions is given to improve the data collection strategy.

AB - Hyper-resolution datasets for urban flooding are rare. This problem prevents detailed flooding risk analysis, urban flooding control, and the validation of hyper-resolution numerical models. We employed social media and crowdsourcing data to address this issue. Natural Language Processing and Computer Vision techniques are applied to the data collected from Twitter and MyCoast (a crowdsourcing app). We found these big data based flood monitoring approaches can complement the existing means of flood data collection. The extracted information is validated against precipitation data and road closure reports to examine the data quality. The two data collection approaches are compared and the two data mining methods are discussed. A series of suggestions is given to improve the data collection strategy.

U2 - 10.1016/j.cageo.2017.11.008

DO - 10.1016/j.cageo.2017.11.008

M3 - Article

AN - SCOPUS:85033720335

VL - 111

SP - 139

EP - 147

JO - Computers and Geosciences

JF - Computers and Geosciences

SN - 0098-3004

ER -