Qilian Mountain National Park (QLMNP) is a biodiversity hotspot with great agriculture and tourism resources. With the expansion of human activities, a few areas of the park are experiencing massive landscape transformation, and these areas are also highly ecologically sensitive. Nevertheless, there are substantial differences in the human activities and natural resources of various communities around QLMNP, resulting in heterogeneous landscape degradation. Hence, this study explores the extent and drivers of spatial heterogeneity in landscape fragmentation associated with ecologically vulnerable communities in QLMNP. Multiple ring buffer analysis and geographically weighted regression (GWR) were used to analyze the relationships between landscape fragmentation and variables of human activities and facilities to identify the main factors influencing landscape fragmentation in different regions. The results reveal that human disturbance had a stronger relationship with landscape fragmentation in QLMNP than natural factors do. Among the drivers of landscape fragmentation, the distribution of residential areas and the extension of agricultural land were found to have more significant impacts than tourism. Expansion of cropland had a greater impact on the eastern part of the national park, where overgrazing and farming require further regulation, while tourism affected the landscape fragmentation in the central area of the national park. The point-shaped human disturbance had a larger impact than the linear disturbance. The study findings can be used to formulate a comprehensive plan to determine the extent to which agriculture and tourism should be developed to avoid excessive damage to the ecosystem.
Under the combined influence of climate change and urban development, the risk of urban flooding caused by extreme weather events has increased significantly, making assessing flood vulnerability and resilience increasingly crucial for urban flood management. With the 45 counties in Qinghai Province as the research objects, the hazard risk of flood and exposure are combined to study their vulnerability. At the same time, resilience is evaluated by the indicators selected from four dimensions (society, economy, environment, and infrastructure). Through Z-scoring, the vulnerability and resilience of each county are clustered into four groups to explore their associations from a spatial balance perspective. Obstacle factor analysis is introduced to summarize the key factors affecting the improvement of urban resilience in Qinghai Provence. The results show that the eastern areas of Qinghai experience high vulnerability to flooding because of high levels of hazard and exposure. What is more, Xining, Haidong, and Haixi experience a high level of resilience. A strong spatial mismatch between vulnerability and resilience exists in Qinghai, with 24 counties (58%) being self-adapted, 8 counties (18%) over-abundant, and 11 counties deficient in terms of nature–nurture. The length of levee and number of beds in medical institutions are the main obstacles to resilience in Qinghai. The research results can provide a theoretical and scientific basis for future urban flood management and resilience development in the Qinghai–Tibetan Plateau.