A new model improves landslide susceptibility assessments during extreme rainfall by examining the interactions between slope units within watersheds.
Recent research highlights growing challenges posed by rainfall-induced landslides, which have become more frequent due to climate change. Extreme weather events not only contribute to devastating landslides but also complicate their prediction and management. Addressing these issues, scientists have developed the Network-based Landslide Susceptibility Assessment Model (NLSAM), which takes hydrological interactions between slope units within watersheds during extreme rainfall events.
The study was led by Chenlu Wang and colleagues, who assessed the model's efficacy in Fuyang District, Zhejiang Province, China. The region is prone to heavy rainfalls due to its geographical location and has experienced significant landslide damage. On July 22, 2023, Fuyang was particularly affected by severe rainfall, which triggered numerous shallow landslides across the mountainous terrain.
Traditional landslide susceptibility models often evaluate slope units individually, failing to recognize the interconnections between these units. This isolation can overlook the reality of landslides occurring in groups, leading to potential miscalculations of risk. NLSAM offers improved functionality by using complex network theory to model water flow between slopes, illustrating how rainfall events propagate hydrologically across the terrain.
The introduction of NLSAM incorporates key physical models, including the Fast Shallow Landslide Assessment Model (FSLAM), to assess changes in slope stability under varying rainfall intensities. The research emphasizes the importance of evaluating both vertical and lateral water infiltration during extreme weather conditions.
Through experimental applications, NLSAM demonstrated precision by identifying instability among slope units more effectively than traditional models. During tested extreme rainfall scenarios, the model showed remarkable accuracy, illustrating how water transfers destabilized more slopes and heightened overall landslide susceptibility. With validation results reflecting NLSAM's recall rate of 0.93, the model proved its adeptness at effectively predicting group-occurring landslides.
Co-author Jianlin Zhou noted, "NLSAM captures rainfall propagation paths and quantifies their impacts, assisting decision-makers in formulating more effective management strategies." Such developments are particularly encouraging for enhancing localized disaster preparedness and response initiatives.
Further insights from the study revealed the dynamic relationship between slope units: as rainfall intensity increases, the hydrological connections become increasingly significant, influencing landslide risk more than previously understood. It was found, "The closer the hydrological connection of the slope unit, the more prone to landslides at the node."
By leveraging complex network frameworks, the research team has provided fresh perspectives on predicting and managing landslide susceptibility under extreme rainfall conditions. This approach has strong potential for future studies aiming to bolster the resilience of vulnerable regions to hydrometeorological hazards.
Conclusions drawn from this study advocate for incorporating networked hydrological relationships for enhanced landslide risk assessments. The enhanced capability of NLSAM to detect cascading failures among slope units not only fills gaps left by traditional models but also promises substantial advancements for disaster risk reduction efforts.