Based on the hourly meteorological datasets from Beijing, China Automatic Weather Stations in 2016, selecting Oct. 23 as the research period without the influence of precipitation and wind, we generated 24 continuous temperature surfaces through ANUSPLIN interpolation software and defined the UHI (urban heat island) using percentile method. The 24-hour moving path and distribution pattern of UHI were analyzed. The relationship was discussed between the diurnal variations of the UHI and the percentage structure of land cover in Beijing. The results show that, 1) there is a diurnal variation of UHI intensity in autumn sunny days in Beijing, and the maximum growth rate is 0.48oC/h at 18:00–23:00. 2) There are three classifications of UHI’s horizontal distribution: single-center UHI, multi-center UHI and radial UHI. 3) The proportion of six types of land use is different in the hourly UHI. In the day, there is no representative land use type in the hourly UHI, and the percentage structure of the six types of land use in the hourly UHI is different. At night, the main land use type in the hourly UHI is built-up land, and the percentage structure of the six types of land use in the hourly UHI is similar. 4) The relationship between the hourly change of UHI and land use type: from 0:00 to 3:00, the area of the hourly moving region of UHI where the proportion of six land use types did not change significantly is small; from 3:00 to 7:00, the UHI moves to the region where the proportion of built-up land decreases and the proportion of forest land increases; from 7:00 to 9:00, the UHI moves to the region where the proportion of forest land decreases and the proportion of built-up land and crop land decreases; from 9:00 to 15:00, the UHI moves to the region where the proportion of built-up land decreases and the proportion of crop land increases; from 15:00 to 19:00, the UHI returnes to the region where the proportion of built-up land increases; from 19:00 to 22:00, the proportion of six land use types in the hourly UHI’s moving region does not change significantly; from 22:00 to 23:00, the UHI moves to the area where the proportion of crop land increases and the proportion of built-up land decreases.

利用2016年北京市自动观测站的逐时气象要素数据, 排除复杂天气影响, 筛选出微风、晴朗的10月23日全天为研究时段。采用ANUSPLIN空间化插值软件和百分位法定义城市热岛, 探究北京市热岛24小时移动路径与空间分布格局特征, 分析热岛逐小时变化与土地覆被的关系, 得到如下结果。1) 北京市秋季晴天热岛强度存在日变化规律, 其中18:00—23:00时段的热岛强度增长速率最大为0.48oC/h。2) 日内不同时刻热岛分布存在差异, 水平分布形态有以下 3 种范式: 单中心式热岛、多中心式热岛和辐射状热岛。3) 不同时刻热岛分布区内6种土地利用类型的面积占比不同, 白天热岛分布区无代表性土地利用类型, 不同时刻热岛区6种土地利用类型的百分比结构差异大; 夜晚热岛分布区土地利用类型以建设用地为主, 不同时刻热岛区 6种土地利用类型的百分比结构差异小。4) 热岛日内变化与土地利用的关系: 00:00—03:00, 热岛逐时移动面积小, 且移动区6种土地利用类型的占比无明显变化; 03:00—07:00, 热岛向林地占比增多、建设用地占比减小的区域移动; 07:00—09:00, 热岛向林地占比减小、耕地与建设用地占比增大的区域移动; 09:00—15:00, 热岛整体向东移动, 建设用地占比减小、耕地占比增大; 15:00—19:00, 热岛移回建设用地占比增大的区域; 19:00—22:00, 热岛移动区内6种土地利用类型的占比无明显变化; 22:00—23:00, 热岛向耕地占比增大、建设用地占比减小的区域移动。