Chia-Fu Liu , Lipai Huang , Kai Yin , Sam Brody , Ali Mostafavi
{"title":"FloodDamageCast:利用机器学习和数据增强技术建立洪水灾害预报系统","authors":"Chia-Fu Liu , Lipai Huang , Kai Yin , Sam Brody , Ali Mostafavi","doi":"10.1016/j.ijdrr.2024.104971","DOIUrl":null,"url":null,"abstract":"<div><div>Near-real-time estimation of damages (a.k.a, damage nowcasting) to building and infrastructure is crucial during response and recovery efforts. Despite advancements in flood risk predictions, the majority of existing methods primarily focus on inundation estimation with limited damage nowcasting capabilities. Flooding damage nowcasting at fine spatial resolutions remains a very challenging problem with currently no existing model to perform the task. This limitation is mainly due to a number of technical challenges such as limited consideration of non-linear interactions between flood hazards and build-environment features, issues with imbalanced datasets, and the absence of reliable ground truth for model performance evaluation. To address this important gap, this study presents FloodDamageCast, a machine learning (ML) framework tailored for property flood damage nowcasting. The framework leverages heterogeneous data related to the built environment, topographic, and hydrological features to predict residential flood damage in a fine resolution of 500 m by 500 m in the context of Harris County, TX, during the 2017 Hurricane Harvey. To deal with data imbalance, FloodDamageCast includes a tabular data augmentation model based on Conditional Tabular Generative Adversarial Networks (CTGAN). The data augmentation model component addresses highly imbalanced class issues, where the majority class constitutes 96.4% of the dataset, potentially impairing model performance, By combining GAN-based data augmentation with an efficient ML model, Light Gradient-Boosting Machine (LightGBM), our results demonstrate the framework’s ability to identify high-damage spatial areas that would be overlooked by baseline models. the satisfactory performance of FloodDamageCast also shows its capability to be used for flood damage nowcasting at a fine spatial resolution to inform response and recovery efforts. The insights from flood damage nowcasting would help emergency management agencies and public officials to more efficiently identify repair needs and allocate resources, and also save time and efforts during on-the-ground inspections.</div></div>","PeriodicalId":13915,"journal":{"name":"International journal of disaster risk reduction","volume":"114 ","pages":"Article 104971"},"PeriodicalIF":4.2000,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"FloodDamageCast: Building flood damage nowcasting with machine-learning and data augmentation\",\"authors\":\"Chia-Fu Liu , Lipai Huang , Kai Yin , Sam Brody , Ali Mostafavi\",\"doi\":\"10.1016/j.ijdrr.2024.104971\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>Near-real-time estimation of damages (a.k.a, damage nowcasting) to building and infrastructure is crucial during response and recovery efforts. Despite advancements in flood risk predictions, the majority of existing methods primarily focus on inundation estimation with limited damage nowcasting capabilities. Flooding damage nowcasting at fine spatial resolutions remains a very challenging problem with currently no existing model to perform the task. This limitation is mainly due to a number of technical challenges such as limited consideration of non-linear interactions between flood hazards and build-environment features, issues with imbalanced datasets, and the absence of reliable ground truth for model performance evaluation. To address this important gap, this study presents FloodDamageCast, a machine learning (ML) framework tailored for property flood damage nowcasting. The framework leverages heterogeneous data related to the built environment, topographic, and hydrological features to predict residential flood damage in a fine resolution of 500 m by 500 m in the context of Harris County, TX, during the 2017 Hurricane Harvey. To deal with data imbalance, FloodDamageCast includes a tabular data augmentation model based on Conditional Tabular Generative Adversarial Networks (CTGAN). The data augmentation model component addresses highly imbalanced class issues, where the majority class constitutes 96.4% of the dataset, potentially impairing model performance, By combining GAN-based data augmentation with an efficient ML model, Light Gradient-Boosting Machine (LightGBM), our results demonstrate the framework’s ability to identify high-damage spatial areas that would be overlooked by baseline models. the satisfactory performance of FloodDamageCast also shows its capability to be used for flood damage nowcasting at a fine spatial resolution to inform response and recovery efforts. The insights from flood damage nowcasting would help emergency management agencies and public officials to more efficiently identify repair needs and allocate resources, and also save time and efforts during on-the-ground inspections.</div></div>\",\"PeriodicalId\":13915,\"journal\":{\"name\":\"International journal of disaster risk reduction\",\"volume\":\"114 \",\"pages\":\"Article 104971\"},\"PeriodicalIF\":4.2000,\"publicationDate\":\"2024-11-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"International journal of disaster risk reduction\",\"FirstCategoryId\":\"89\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S2212420924007337\",\"RegionNum\":1,\"RegionCategory\":\"地球科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"GEOSCIENCES, MULTIDISCIPLINARY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"International journal of disaster risk reduction","FirstCategoryId":"89","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2212420924007337","RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"GEOSCIENCES, MULTIDISCIPLINARY","Score":null,"Total":0}
FloodDamageCast: Building flood damage nowcasting with machine-learning and data augmentation
Near-real-time estimation of damages (a.k.a, damage nowcasting) to building and infrastructure is crucial during response and recovery efforts. Despite advancements in flood risk predictions, the majority of existing methods primarily focus on inundation estimation with limited damage nowcasting capabilities. Flooding damage nowcasting at fine spatial resolutions remains a very challenging problem with currently no existing model to perform the task. This limitation is mainly due to a number of technical challenges such as limited consideration of non-linear interactions between flood hazards and build-environment features, issues with imbalanced datasets, and the absence of reliable ground truth for model performance evaluation. To address this important gap, this study presents FloodDamageCast, a machine learning (ML) framework tailored for property flood damage nowcasting. The framework leverages heterogeneous data related to the built environment, topographic, and hydrological features to predict residential flood damage in a fine resolution of 500 m by 500 m in the context of Harris County, TX, during the 2017 Hurricane Harvey. To deal with data imbalance, FloodDamageCast includes a tabular data augmentation model based on Conditional Tabular Generative Adversarial Networks (CTGAN). The data augmentation model component addresses highly imbalanced class issues, where the majority class constitutes 96.4% of the dataset, potentially impairing model performance, By combining GAN-based data augmentation with an efficient ML model, Light Gradient-Boosting Machine (LightGBM), our results demonstrate the framework’s ability to identify high-damage spatial areas that would be overlooked by baseline models. the satisfactory performance of FloodDamageCast also shows its capability to be used for flood damage nowcasting at a fine spatial resolution to inform response and recovery efforts. The insights from flood damage nowcasting would help emergency management agencies and public officials to more efficiently identify repair needs and allocate resources, and also save time and efforts during on-the-ground inspections.
期刊介绍:
The International Journal of Disaster Risk Reduction (IJDRR) is the journal for researchers, policymakers and practitioners across diverse disciplines: earth sciences and their implications; environmental sciences; engineering; urban studies; geography; and the social sciences. IJDRR publishes fundamental and applied research, critical reviews, policy papers and case studies with a particular focus on multi-disciplinary research that aims to reduce the impact of natural, technological, social and intentional disasters. IJDRR stimulates exchange of ideas and knowledge transfer on disaster research, mitigation, adaptation, prevention and risk reduction at all geographical scales: local, national and international.
Key topics:-
-multifaceted disaster and cascading disasters
-the development of disaster risk reduction strategies and techniques
-discussion and development of effective warning and educational systems for risk management at all levels
-disasters associated with climate change
-vulnerability analysis and vulnerability trends
-emerging risks
-resilience against disasters.
The journal particularly encourages papers that approach risk from a multi-disciplinary perspective.