科研速览继续刷下去 →
◆ Journal of environmental management2026-09-26

Operationalizing an ecogeomorphological framework for national-scale landslide susceptibility: A hybrid PCA-AHP approach to support risk-informed spatial planning in Austria.

Zhiyi Zhang, Anna M Hersperger, Michael U Hensel

一句话结论

These findings show that broader predictor coverage does not necessarily translate into better overall discrimination, but that additional ecogeomorphological information can provide distinct analytical value when evaluated systematically.

原始摘要(原文)
Human settlements, urban centers, and critical infrastructure in Alpine regions are increasingly exposed to natural hazards. In Austria, while disaster risk management is coordinated nationally, institutional governance is currently distributed across nine federal provinces, resulting in a patchwork of regional datasets and mapping practices. To support greater consistency at the national scale, this study proposes an ecogeomorphological conceptual framework (EGCF) for risk-informed spatial planning, structured around three thematic pillars: ecology, geomorphology, and hydrology. The framework is operationalized through spatial multi-criteria evaluation (SMCE), combining 1) development-only frequency ratio (FR) scores with 2) principal component analysis (PCA)-based within-theme predictor weights and 3) analytic hierarchy process (AHP) theme weights. It is implemented for national-scale landslide susceptibility assessment in Austria using a 17-factor EGCF-informed configuration and a literature-derived 9-factor baseline evaluated within the same workflow. The resulting 17-factor landslide susceptibility index covered 89.88% of national grid cells. In the primary spatial holdout evaluation, the full 17-factor configuration did not outperform the 9-factor baseline; the area under the curve (AUC) values were 0.787 and 0.806 for the 17-factor and 9-factor configurations, respectively. However, development-only factor-addition diagnostics showed that the contribution of the added predictors was heterogeneous, with additional hydrological predictors providing the most consistent improvement across five spatial folds. These findings show that broader predictor coverage does not necessarily translate into better overall discrimination, but that additional ecogeomorphological information can provide distinct analytical value when evaluated systematically. The framework therefore provides a reproducible national susceptibility baseline and a transparent structure for evaluating the contribution of additional ecogeomorphological information. This supports national-scale screening and risk-informed spatial planning.
读原文 ↗

AI 追问PRO

登录后使用 AI 追问

讨论区

登录后参与讨论

相关论文

Operationalizing an ecogeomorphological framework for national-scale landslide susceptibility: A hybrid PCA-AHP approach to support risk-informed spatial planning in Austria. — 科研速览 Science Skim