Abstract:

The exploration of borderland land use patterns and trends is crucial for maintaining borderland security and promoting high-quality border development. This study employed an improved K-means clustering algorithm for spatial geographic zoning and utilized methods such as kernel density analysis and Geodetectors to analyze trends in land use change in the Guangxi section of the China-Vietnam border. Examining current border security issues, two scenarios were set up for land use simulation and prediction: natural development and national security. The results indicate that: ①The simulation accuracy significantly increased after spatial geographic zoning, and the conditions set for the simulation align with the research needs of national security strategic development. ②Under the natural development scenario, phenomena such as land abandonment and destructive planting continue to exacerbate the vacuity along the border. Under the national security scenario, construction land maintains a higher growth rate, and the trend of reducing permanent basic farmland and medium to high canopy density forest areas is contained. ③The differences in land use between the two scenarios are mainly reflected in three types of land: farmland, medium to high canopy forest land, and construction land. The contradictions between social change, economic development demands, and land use methods have a subtle impact on the security of the Guangxi section of the China-Vietnam border. ④In the future, land use in the Guangxi section of the China-Vietnam border should be guided by national security strategies, through implementing differentiated spatial controls of borderland use and in-situ urbanization of border villages, leading to safe and stable development of the border area.

Key words: land management, national security, cellular automaton, spatial clustering, Guangxi section of the China-Vietnam border

摘要：

探索边境土地利用规律和趋势对维护边境土地安全和推进边境高质量发展意义重大。采用改进的K均值聚类算法进行空间地理分区，运用核密度分析、地理探测器等方法分析中越边境广西段的用地变化趋势。通过审视现有边境安全问题，设置自然发展和国家安全两类情景来模拟预测研究区未来的土地利用变化，结果表明：①空间地理分区后的模拟精度显著提高，模拟设置的条件参数符合国家安全战略发展的研究需求。②在自然发展情景下，不合理用地结构持续加剧边境虚空态势。在国家安全情景下，建设用地保持较高增长速度，永久基本农田、中高郁闭度林地减少趋势得到遏制。③两种情景下的土地利用差异主要体现在耕地、中高郁闭度林地、建设用地三类用地上，社会变迁、经济发展需求与土地利用方式之间的矛盾影响着中越边境广西段的安全。④未来中越边境广西段的土地利用应以国家安全战略为指引，通过实施差异化的边境国土空间管控、沿边村庄原地城镇化等措施来引领边境地区安全稳定发展。

关键词: 土地管理, 国家安全, 元胞自动机, 空间聚类, 中越边境广西段