Scholarlink

<p>Flooding is a recurring hazard in Greater Bandung that requires time-sensitive hazard assessment for effective mitigation. This study develop a spatio-temporal flood-hazard mapping framework by applying survival analysis within a GIS environment. The study models flood occurrence timing and spatial hazard using the Kaplan–Meier estimator and the Cox Proportional Hazards model. Eight conditioning factors are used: TWI, elevation, slope, rainfall, land cover, NDVI, distance from the river, and distance from the road. Model training and validation used historical flood records (2015–2022) and time-dependent ROC/AUC. The Cox model achieved excellent predictive accuracy (AUC = 0.927). Rainfall (HR ≈ 1.434) and TWI (HR ≈ 1.164) increased hazard, while NDVI strongly reduced it (HR ≈ 0.03). Hazard maps identify northern and central Greater Bandung (Bandung and Cimahi urban cores) as highest hazard areas. The time-aware survival-analysis approach supplies planners with spatial hazard maps and probabilistic timing for targeted mitigation and early warning.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Dynamic flood hazard mapping in greater Bandung using the Cox proportional hazards model

  • 期刊论文

  • 发表日期：

    2025年12月26日

查看全文
  • Riki Purnama Putra,
  • Riantini Virtriana,
  • Agustinus Bambang Setyadji,
  • Rena Denya Agustina

摘要

Flooding is a recurring hazard in Greater Bandung that requires time-sensitive hazard assessment for effective mitigation. This study develop a spatio-temporal flood-hazard mapping framework by applying survival analysis within a GIS environment. The study models flood occurrence timing and spatial hazard using the Kaplan–Meier estimator and the Cox Proportional Hazards model. Eight conditioning factors are used: TWI, elevation, slope, rainfall, land cover, NDVI, distance from the river, and distance from the road. Model training and validation used historical flood records (2015–2022) and time-dependent ROC/AUC. The Cox model achieved excellent predictive accuracy (AUC = 0.927). Rainfall (HR ≈ 1.434) and TWI (HR ≈ 1.164) increased hazard, while NDVI strongly reduced it (HR ≈ 0.03). Hazard maps identify northern and central Greater Bandung (Bandung and Cimahi urban cores) as highest hazard areas. The time-aware survival-analysis approach supplies planners with spatial hazard maps and probabilistic timing for targeted mitigation and early warning.