本文以巴音布鲁克世界自然遗产地为研究对象，基于卫星遥感影像、气象站点观测信息、地面调查资料等多源数据，完成了1992年—2020年遗产地突出普遍价值OUVs（Outstanding Universal Values）关键要素空间制图，并从景观尺度分析OUVs关键要素及景观格局的时空变化特征，讨论影响其变化特征的主要驱动因子。结果表明：遗产地从中心向边缘呈水体—沼泽—高覆盖度草地—中覆盖度草地/滩地/沙地—低覆盖度草地/耕地—裸地/建设用地逐渐过渡的景观特征。1992年—2020年遗产地以沼泽和中覆盖度草地为主要OUVs表征要素，其年平均面积分别占遗产地总面积的29.51%和29.75%。整个遗产地景观则呈散布且复杂化分布特征，斑块密度（PD）、形状指数（LSI）由1992年的2.05个/km²、43.01分别增加到2020年的2.37个/km²、52.02，蔓延度指数（CONTAG）则由58.44降低为56.54。研究发现遗产地水体和沼泽受平均降水量影响最大，r绝对值为0.884、0.929，低覆盖度草地受平均相对湿度影响最大，r绝对值为0.931，说明自然因素是遗产地高寒湿地和高寒草地面积变化的主要因子。人口密度、土壤类型和高程为影响遗产地OUVs关键要素空间分异特征的主要因子，解释力在30%—45%；各因子的交互作用均高于单因子的影响，人口密度与土壤类型组合的解释力最大达47.8%；驱动因子对遗产地OUVs关键要素动态度的影响程度则均不高（最大q值仅为0.106）。巴音布鲁克自然遗产地景观格局变化及其驱动因素研究结果对高山湿地类自然遗产地保护、生态修复及可持续发展等具有重要意义。

The World Natural Heritage Sites has attracted worldwide focus because of their Outstanding Universal Values (OUVs). However, in recent years, fragile ecological environments have destroyed the structural and functional integrity of the ecological system of the heritage sites, seriously threatening the OUVS of the World Natural Heritage Sites. In this paper, on the basis of satellite remote sensing image, meteorological observation data, and ground survey data, taking Bayanbulak world natural heritage as a study case, we mapped the spatial key elements of OUVs in Bayanbulak heritage from 1992 to 2020. Combined with a dynamic degree and landscape pattern index, the temporal and spatial change characteristics of OUVs key elements and their landscape pattern were analyzed, and the main natural and human factors influencing their spatial distribution and change characteristics were discussed. Results show that a landscape characteristic of gradual transition of water-marsh-high coverage grassland-moderate coverage grassland/beach land/sandy land-low coverage grassland/farmland-barren/construction land was presented from the center of the heritage protection area to the edge. Moreover, the main representative elements were marsh and moderate coverage grassland, which accounted for 29.51% and 29.75% of the study area, respectively. Concerning the landscape pattern, it shows scattered and complicated development characteristics. PD and LSI increased from 2.05 and 43.01 in 1992 to 2.37 and 52.02 in 2020. The contagion index has also been reduced from 58.44% to 56.54%. The average precipitation was the key factor influencing the area of water and marsh, and their r values are 0.884 and -0.929, respectively. The area of low coverage grassland was most affected by the average relative humidity, with r value of -0.931. These indicate that natural factors were the main reasons for the area change of alpine wetland and alpine grassland in Bayanbulak. Population density, soil type, and elevation were the main factors driving the spatial distribution of OUVs key elements in Bayanbulak Heritage Site, and the maximum q value in Geodetector was 0.45. The interaction effect between any two factors was greater than that of individual factor, and the maximum detector power of 47.8% was in the combination of population density and soil type. However, the detector powers of factors on the dynamic degree of OUVs key elements in Bayanbuk site were not high (the maximum q value was only 0.106). The study of landscape pattern variation and its driving factors in natural heritage sites are highly significant for heritage protection, ecological restoration, and sustainable development.