Home >

news Help

Publication Information


Title
Japanese: 
English:Building information modeling integrated with environmental flood hazard to assess the building vulnerability to flash floods 
Author
Japanese: Mohamed Wahba, Mahmoud Sharaan, Wael M Elsadek, 鼎 信次郎, H Shokry Hassan.  
English: Mohamed Wahba, Mahmoud Sharaan, Wael M Elsadek, Shinjiro Kanae, H Shokry Hassan.  
Language English 
Journal/Book name
Japanese: 
English:Stochastic Environmental Research and Risk Assessment 
Volume, Number, Page         1-21
Published date Jan. 13, 2024 
Publisher
Japanese: 
English:Springer Berlin Heidelberg 
Conference name
Japanese: 
English: 
Conference site
Japanese: 
English: 
Official URL https://link.springer.com/article/10.1007/s00477-023-02640-9
 
DOI https://doi.org/10.1007/s00477-023-02640-9
Abstract Recently, significant global regions have encountered unparalleled consequences arising from climate change, notably manifesting in the heightened intensity of flash floods which presents a formidable and complex challenge for numerous governments globally. The main goal of this research is quantifying the Building Vulnerability Index (BVI), a metric devised to assess the susceptibility of buildings to flash floods. This study introduces a two-faceted index focusing on Intrinsic Vulnerability (IV) and Environmental Flood Hazard (EFH). The IV pertains to intrinsic characteristics of structures that influence their ability to withstand and recover from flash floods, whereas “EFH” involves the evaluation of potential hazards from external flooding events through the utilization of eleven environmental parameters. This assessment employs machine learning, specifically an Artificial Neural Network with Multilayer Perceptron architecture, functioning as both classifier and regressor models. The EFH classifies 40,521 structures according to five levels of vulnerability, with level one indicating a very low hazard and level 5 signifying a very high hazard. The majority of buildings, approximately 90%, fall into the categories of levels 2 and 3, indicating low and moderate flood hazard. Receiver Operating Characteristic (ROC) analysis validates high accuracy of the two adopted machine learning approaches after estimating area under curve as 94.27% and 95.22% for classifier and regressor models respectively. Consequentially, calculation of BVI across 36 structures underscores a spectrum of susceptibility, ranging from 0.471 to 0.795. Notably, this investigation introduces a novel approach by amalgamating building-specific data with environmental flood hazard assessment, thereby enhancing the evaluation of flood vulnerability. Finally, the study successfully estimates integrated flood vulnerability and offers valuable insights for decision-makers in devising mitigation strategies.

©2007 Tokyo Institute of Technology All rights reserved.