高寒山地生态脆弱区聚落空间格局特征及成因识别——以天祝藏族自治县为例

自然环境复杂多样是山区聚落空间差异性和异质性的主要原因,对聚落空间治理及可持续发展具有深刻影响。采用天祝县2018年土地变更调查数据,利用空间"热点"探测、最小累积阻力值模型（MCR）和二元Logistic回归等方法,对聚落空间分布格局及其形成因素进行分析。结果表明：（1）聚落分布整体表现为集聚型,空间上呈现出"东南高、西北低"的格局。（2）核密度估计值东南高、西北低,形成"东南密集型"和"西北稀疏型"2个典型分布区,呈现出"核心-边缘"结构;同时,聚落规模空间"热点"探测显示,聚落形态具有显著的空间差异性,华藏寺镇表现为高密度大斑块,其他乡镇表现为中低密度小斑块。（3）聚落斑块平均密度对分组分析结果作用显著,聚落类型由中低密度斑块主导,受高寒气候区分布范围较广的影响,高寒区稀疏型聚落覆盖范围较大。（4）通过聚落空间格局影响因素统计分析和Logistic回归分析,定量识别出：地形条件、土地资源配置和降水条件对聚落分布疏密程度作用最显著,交通禀赋和交通可达性对县域聚落格局优化具有重要作用。

Spatial pattern characteristics and genetic identification of settlements in ecologically fragile areas of alpine mountains: a case study on the Tianzhu Tibetan Autonomous County

The complex and diverse natural environment is the main reason for the spatial difference and heterogeneity of settlements in mountainous areas, which has a profound impact on the spatial governance and sustainable development of settlements. According to the land change survey data of Tianzhu county in 2018, the spatial distribution pattern and its forming factors of settlements were analyzed by the spatial "hot spot" detection, the minimum cumulative resistance value model (MCR) and binary logistic regression. The results show that: (1) the overall distribution of settlements is agglomeration. And it is high in the southeast and low in the northwest in the spatial pattern. (2) The kernel density estimate of settlement patches is high in the southeast and low in the northwest, which causes the formation of two typical distribution areas: "southeast intensive" and "northwest sparse". This phenomenon presents a "core-edge" structure. Meanwhile, settlement "hot spot" detection shows that there is a significant spatial difference in settlement patterns. Huazangsi town showed large patches of high density, while other towns showed small patches of medium and low density. (3) The average density of settlement patches has a significant effect on the results of grouping analysis. The settlement type is dominated by medium and low density patches. And it is affected by the wide distribution range of the alpine climate zone and the sparse settlement in the alpine region covers a large area. (4) The influencing factors of settlement spatial pattern were analyzed by statistical analysis and logistic regression analysis. We identified quantitatively that topographic conditions, land resource allocation and precipitation conditions had the most significant effect on the density of settlement distribution. Transportation endowment and accessibility play a crucial role in the optimization of settlement pattern in county areas.