Landslide susceptibility assessment is an important means of helping to reduce and manage landslide risk. The existing studies, however, fail to examine the spatially varying relationships between landslide susceptibility and its explanatory factors. This paper investigates the spatial variation in such relationships in Liangshan, China, leveraging a spatially explicit model, namely, geographical random forest (GRF). By comparing with random forest (RF), we found that GRF achieves a higher performance with an AUC of 0.86 due to its consideration of the spatial heterogeneity among variables. GRF also provides a higher-quality landslide susceptibility map than RF by correctly placing 92.35% of the landslide points in high-susceptibility areas. The local feature importance derived from GRF allows us to understand that the impact of conditioning factors varies across space, which can provide implications for policy development by local governments to place different levels of attention on different conditioning factors in specific counties to prevent and mitigate landslides. To account for the spatial dependence among the data in the model performance assessment, we use spatial cross-validation (CV) to split the data into subsets spatially rather than randomly for model training and testing. The results show that spatial CV can effectively address the over-optimistic bias in model error evaluation.

Despite the importance of health vulnerability in disaster risk assessment, most of the existing disaster vulnerability indicators only emphasize economic and social vulnerability. Important underlying health risks such as non-communicable disease are not included in vulnerability measures. A three-phase methodology approach was used to construct a disaster risk model that includes a number of key health indicators which might be missing in global disaster risk analysis. This study describes the development of an integrated health vulnerability index and explains how the proposed vulnerability index may be incorporated into an all-hazard based disaster risk index in the Belt and Road Initiative (BRI), also known as the “Silk Road Economic Belt”, region. Relevant indicators were identified and reviewed in the published literature in PubMed/Medline. A two-stage dimension reduction statistical method was used to determine the weightings of relevant dimensions to the construction of the overall vulnerability index. The proposed final health vulnerability index included nine indicators, including the proportion of the population below 15 and above 65 years, under-five mortality ratio, maternal mortality ratio, tuberculosis prevalence, age-standardized raised blood pressure, physician ratio, hospital bed ratio, and coverage of the measles-containing-vaccine first-dose (MCV1) and diphtheria tetanus toxoid and pertussis (DTP3) vaccines. This proposed index, which has a better reflection of the health vulnerability in communities, may serve as a policy and implementation tool to facilitate the capacity-building of Health-Emergency Disaster Risk management (Health-EDRM).

This study aimed to develop a physical-based approach for predicting the spatial likelihood of shallow landslides at the regional scale in a transition zone with extreme topography. Shallow landslide susceptibility study in an area with diverse vegetation types as well as distinctive geographic factors (such as steep terrain, fractured rocks, and joints) that dominate the occurrence of shallow landslides is challenging. This article presents a novel methodology for comprehensively assessing shallow landslide susceptibility, taking into account both the positive and negative impacts of plants. This includes considering the positive effects of vegetation canopy interception and plant root reinforcement, as well as the negative effects of plant gravity loading and preferential flow of root systems. This approach was applied to simulate the regional-scale shallow landslide susceptibility in the Dadu River Basin, a transition zone with rapidly changing terrain, uplifting from the Sichuan Plain to the Qinghai–Tibet Plateau. The research findings suggest that: (1) The proposed methodology is effective and capable of assessing shallow landslide susceptibility in the study area; (2) the proposed model performs better than the traditional pseudo-static analysis method (TPSA) model, with 9.93% higher accuracy and 5.59% higher area under the curve; and (3) when the ratio of vegetation weight loads to unstable soil mass weight is high, an increase in vegetation biomass tends to be advantageous for slope stability. The study also mapped the spatial distribution of shallow landslide susceptibility in the study area, which can be used in disaster prevention, mitigation, and risk management.