推进阿拉巴马州沿海住宅物业的洪水损害建模:一种多变量机器学习方法
Mujungu Lawrence Museru1, Rouzbeh Nazari2, Abolfazl N Giglou1
1Sustainable Smart Cities Research Center, University of Alabama at Birmingham (UAB), Birmingham, AL, USA; Department of Civil, Construction, and Environmental Engineering, University of Alabama-Birmingham, Birmingham, AL 35294-4440, USA.
The Science of the total environment
|October 18, 2023
概括
准确的洪水风险预测至关重要. 这项研究使用机器学习和过量采样技术增强了洪水损害评估模型,提高了受损物业和相对洪水损害的预测准确度.
科学领域:
- 环境科学 环境科学
- 数据科学数据科学数据科学
- 土木工程 土木工程是指土木工程.
背景情况:
- 洪水构成了全球重大风险,需要准确的洪水风险预测以减轻和提高公众意识.
- 开发强大的洪水损害预测模型受到全面数据集稀缺的阻碍.
- 现有的模型经常与数据不平衡和有限的变量作斗争.
研究的目的:
- 加强国家洪水保险计划 (NFIP) 索赔数据集,用于多变量洪水损害评估.
- 开发和验证先进的机器学习模型,用于预测洪水引起的财产损失.
- 提高沿海地区洪水损害预测的准确性和可靠性.
主要方法:
- 将NFIP索赔数据与阿拉巴马州沿海地区的财产和洪水危险数据结合起来.
- 采用过量抽样技术来解决组合数据集中的数据不平衡问题.
- 利用集体机器学习算法 (随机森林,额外树,XGBoost,CatBoost) 进行模型开发.
主要成果:
- 极端梯度增强 (XGBoost) 在识别损坏的属性方面表现出卓越的性能 (精度:0.89,回忆:0.90,F1分数:0.90).
- XGBoost有效地确定了相对洪水损害 (R平方:0.59,RMSE:0.21,斯皮尔曼相关性:0.70).
- 数据过量采样在不平衡的洪水损害数据集上显著改善了模型性能.
结论:
- 增强的数据集和机器学习,特别是过量采样的XGBoost,为多变量洪水损害评估提供了强大的方法.
- 该研究提供了一种经过验证的方法来提高洪水损害预测的准确性.
- 沙普利添加式解释 (SHAP) 提供了对模型预测的建设性见解,与预期的特征相互作用保持一致.
相关概念视频
Design Example: Analyzing Capacity Contours for Flood Risk Assessment
52
Flood risk assessment involves careful planning and analysis to ensure the safety of communities near water retention structures. Capacity contours are a vital tool in this process, as they illustrate the potential spread of water at specific levels in a given area. In the context of building a bund across a small valley, these contours play a critical role in evaluating the safety of nearby residential areas.In this example, the bund is intended to store stormwater in the valley. The engineers...
52
Design Example: Creating a Hydraulic Model of a Dam Spillway
207
Scaled hydraulic models of dam spillways provide a practical way to replicate and study the intricate flow dynamics of these structures. Often built to a 1:15 ratio, these models allow for observing critical water behavior, such as velocity distribution, flow patterns, and energy dissipation.
207
Modeling and Similitude
271
Scaled modeling is a fundamental technique in engineering, enabling the study of large and complex systems by creating smaller, manageable replicas that recreate critical characteristics of the original. In hydrology and civil infrastructure, for example, scaled models of dams help analyze water flow, turbulence, and pressure. This method allows for accurate predictions of real-world behavior within a controlled environment, significantly reducing the cost and time involved in full-scale...
271
Typical Model Studies
365
Fluid mechanics model studies often utilize scaled-down systems to predict fluid behavior in full-scale environments, such as river flows, dam spillways, and structures interacting with open surfaces. Maintaining Froude number similarity in river models is crucial, as it replicates surface flow features like wave patterns and velocities.
365
Applications of GIS: Disaster Management and Emergency Response
90
Geographic Information System (GIS) technology is essential for risk identification, action prioritization, and resource optimization in critical situations like flooding and earthquakes. By integrating spatial and demographic data, GIS provides a comprehensive framework for emergency response.GIS integrates data layers, like rainfall intensity, topography, elevation profiles, and river levels, to model high-risk flood zones. These layers assess areas susceptible to flooding based on their...
90
Design Example: Maintaining Level of an Embankment
74
Constructing a roadway embankment over uneven terrain requires precise leveling to ensure stability and proper drainage. Surveyors use a leveling instrument and staff to calculate ground elevations and determine the required fill material at each point along the embankment alignment.The process begins by positioning a leveling instrument near a benchmark with a known elevation. A backsight reading establishes the instrument height, which serves as a reference for subsequent measurements. A...
74


