Scholarlink

<p>This study introduces an innovative optimization approach for landslide susceptibility assessment in the central urban area of Badong County, integrating surface deformation rates derived from Small Baseline Subset Interferometric Synthetic Aperture Radar (SBAS-InSAR) technology. To address data gaps caused by SAR image incoherence in certain areas, a neural network-based model was employed to predict deformation rates, treating these as dynamic factors influencing landslide susceptibility. Fourteen static influencing factors were selected to establish a spatial database, complemented by a landslide inventory map containing 86 events. Subsequently, four machine learning models—logistic regression (LR), random forest (RF), support vector machine (SVM), and multilayer perceptron neural network (MLPNN)—were developed to assess landslide susceptibility. Model performance was evaluated using receiver operating characteristic curve (ROC) and area under curve (AUC) metrics. The results revealed that surface deformation rate ranged from -86.98&#xa0;mm/a to 52.73&#xa0;mm/a during the period from January 2019 to January 2021. Incorporating InSAR-derived deformation rates significantly improved the predictive performance of all models, with RF demonstrating the highest accuracy. The generated landslide susceptibility maps (LSMs) revealed that areas with high and very high susceptibility were primarily concentrated along rivers and roads. The model interpretation indicated that InSAR deformation rate (IDR) had an obvious impact on landslide occurrence, particularly when deformation rates exceeded 20 mm/a, although this may introduce certain uncertainties into the landslide susceptibility. This research highlights the significance of incorporating the IDR in enhancing the performance and accuracy of regional landslide susceptibility modelling, providing valuable insights for effective landslide risk management.</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

Enhancing landslide susceptibility modelling through predicted InSAR deformation rates

  • 期刊论文

  • 发表日期：

    2025年6月7日

查看全文
  • Peng Wang,
  • Hongwei Deng,
  • Yanyan Li,
  • Zheng Pan,
  • Tao Peng

摘要

This study introduces an innovative optimization approach for landslide susceptibility assessment in the central urban area of Badong County, integrating surface deformation rates derived from Small Baseline Subset Interferometric Synthetic Aperture Radar (SBAS-InSAR) technology. To address data gaps caused by SAR image incoherence in certain areas, a neural network-based model was employed to predict deformation rates, treating these as dynamic factors influencing landslide susceptibility. Fourteen static influencing factors were selected to establish a spatial database, complemented by a landslide inventory map containing 86 events. Subsequently, four machine learning models—logistic regression (LR), random forest (RF), support vector machine (SVM), and multilayer perceptron neural network (MLPNN)—were developed to assess landslide susceptibility. Model performance was evaluated using receiver operating characteristic curve (ROC) and area under curve (AUC) metrics. The results revealed that surface deformation rate ranged from -86.98 mm/a to 52.73 mm/a during the period from January 2019 to January 2021. Incorporating InSAR-derived deformation rates significantly improved the predictive performance of all models, with RF demonstrating the highest accuracy. The generated landslide susceptibility maps (LSMs) revealed that areas with high and very high susceptibility were primarily concentrated along rivers and roads. The model interpretation indicated that InSAR deformation rate (IDR) had an obvious impact on landslide occurrence, particularly when deformation rates exceeded 20 mm/a, although this may introduce certain uncertainties into the landslide susceptibility. This research highlights the significance of incorporating the IDR in enhancing the performance and accuracy of regional landslide susceptibility modelling, providing valuable insights for effective landslide risk management.