An exploratory study of rural landscape health assessment based on multi-functionality and the integrated weighting method in nine villages of Guangzhou

Rural landscape health assessment, which integrates ecology and culture, is crucial for China’s rural revitalization strategy. However, existing studies primarily focus on large-scale regions, neglecting small-scale areas. This study established a rural landscape health assessment system encompassing ecological conservation, recreational utilization, and socioeconomic production taking nine typical villages in Guangzhou as case studies for exploration. Principal component analysis and the entropy weight method determined indicator weights, and a comprehensive spatial assessment was conducted using a grid-based method. Overall rural landscape health in Guangzhou was relatively low, with the combined area proportion of the “very poor” and “poor” grades > 23.96%; the average area proportion of the “very good” and “good” grades was < 8%, mostly distributed at junctions between patches and corridors with different landscape functions. Significant differences among village types were observed only for the ecological conservation function, with the highest disparity in the area proportion of “very poor” and “poor” grades within villages reaching 48.55%. No significant correlation was found between landscape health and the forest-to-non-forest area ratio; however, a positive correlation with the numerical ratio was identified under ecological conservation and comprehensive functions. The area and numerical ratio maintained a significant positive correlation across all functions. This study reveals the multidimensional characteristics and spatial differentiation of rural landscape health in Guangzhou, providing scientific evidence for targeted optimization. Due to restrictions imposed by urban no-fly zones on drone data collection, this study’s sample only covered nine villages across four districts in Guangzhou. Relying solely on data from the year 2025 and analysis at a single grid scale, it was unable to fully capture spatiotemporal dynamics and multi-scale effects.