Harnessing ML and GIS for Seismic Vulnerability Assessment and Risk Prioritization

被引:0
作者
Shalu [1 ]
Acharya, Twinkle [1 ]
Gharekhan, Dhwanilnath [1 ]
Samal, Anddipak [2 ]
机构
[1] Ctr Environm Planning & Technol CEPT Univ, Fac Technol, Kasturbhai Lalbhai Campus, Ahmadabad 380009, India
[2] Tata Inst Social Sci, VN Purav Marg, Mumbai 400088, India
来源
REVUE INTERNATIONALE DE GEOMATIQUE | 2024年 / 33卷
关键词
Machine learning; earthquake; artificial neural network; random forest; seismic vulnerability; HAZARD; INDIA;
D O I
10.32604/rig.2024.051788
中图分类号
TP7 [遥感技术];
学科分类号
081102 ; 0816 ; 081602 ; 083002 ; 1404 ;
摘要
Seismic vulnerability modeling plays a crucial role in seismic risk assessment, aiding decision-makers in pinpointing areas and structures most prone to earthquake damage. While machine learning (ML) algorithms and Geographic Information Systems (GIS) have emerged as promising tools for seismic vulnerability modeling, there remains a notable gap in comprehensive geospatial studies focused on India. Previous studies in seismic vulnerability modeling have primarily focused on specific regions or countries, often overlooking the unique challenges and characteristics of India. In this study, we introduce a novel approach to seismic vulnerability modeling, leveraging ML and GIS to address these gaps. Employing Artificial Neural Networks (ANN) and Random Forest algorithms, we predict damage intensity values for earthquake events based on various factors such as location, depth, land cover, proximity to major roads, rivers, soil type, population density, and distance from fault lines. A case study in the Satara district of Maharashtra underscores the effectiveness of our model in identifying vulnerable buildings and enhancing seismic risk assessment at a local level. This innovative approach not only fills the gap in existing research by providing predictive modeling for seismic damage intensity but also offers a valuable tool for disaster management and urban planning decision-makers.
引用
收藏
页码:111 / 134
页数:24
相关论文
共 26 条
[1]   Principles and procedures of exploratory data analysis [J].
Behrens, JT .
PSYCHOLOGICAL METHODS, 1997, 2 (02) :131-160
[2]  
Bosher L., 2017, Disaster Risk Reduction for the Built Environment
[3]   Levee failures and social vulnerability in the sacramento-San Joaquin delta area, California [J].
Burton, Christopher ;
Cutter, Susan L. .
2008, American Society of Civil Engineers (ASCE) (09) :136-149
[4]   Recommendations for the quantitative analysis of landslide risk [J].
Corominas, J. ;
van Westen, C. ;
Frattini, P. ;
Cascini, L. ;
Malet, J. -P. ;
Fotopoulou, S. ;
Catani, F. ;
Van Den Eeckhaut, M. ;
Mavrouli, O. ;
Agliardi, F. ;
Pitilakis, K. ;
Winter, M. G. ;
Pastor, M. ;
Ferlisi, S. ;
Tofani, V. ;
Hervas, J. ;
Smith, J. T. .
BULLETIN OF ENGINEERING GEOLOGY AND THE ENVIRONMENT, 2014, 73 (02) :209-263
[5]  
de Los Santos MJ, 2021, Int Arch Photogramm, Remote Sens Spatial Inf Sci, V46, P125
[6]   Seismic Vulnerability Assessment at an Urban Scale by Means of Machine Learning Techniques [J].
Ferranti, Guglielmo ;
Greco, Annalisa ;
Pluchino, Alessandro ;
Rapisarda, Andrea ;
Scibilia, Adriano .
BUILDINGS, 2024, 14 (02)
[7]   An ecologically constrained procedure for sensitivity analysis of Artificial Neural Networks and other empirical models [J].
Franceschini, Simone ;
Tancioni, Lorenzo ;
Lorenzoni, Massimo ;
Mattei, Francesco ;
Scardi, Michele .
PLOS ONE, 2019, 14 (01)
[8]   Two-way interaction of input variables in the sensitivity analysis of neural network models [J].
Gevrey, Muriel ;
Dimopoulos, Ioannis ;
Lek, Sovan .
ECOLOGICAL MODELLING, 2006, 195 (1-2) :43-50
[9]   Delineation of probable seismic sources in India and neighbourhood by a comprehensive analysis of seismotectonic characteristics of the region [J].
Gupta, I. D. .
SOIL DYNAMICS AND EARTHQUAKE ENGINEERING, 2006, 26 (08) :766-790
[10]  
Kuznetsova N., 2014, Random forest visualization