巢湖流域地表热环境与景观变化相关分析及其尺度效应

摘要

基于Landsat ETM+/TIRS影像数据,采用大气校正法反演巢湖流域地表温度,分析2000～2013年的流域地表热环境效应演变规律,通过5个尺度的网格化采样,在不同幅度水平上定量分析LST与下垫面景观的相关性及其尺度效应.结果表明:2000～2013年,巢湖流域LST高强度区域主要以合肥市区为集聚中心并向城市周边蔓延,低强度区域逐渐缩小,流域地表热环境效应增强.LST与下垫面景观结构相关关系分析表明,建设用地景观是流域热环境效应的主要原因,而水体、林地景观可有效抑制地表温度;景观格局对热环境效应影响显著,建设用地与农地景观的破碎度、分离度指数与LST呈极显著负相关,而林地、水体景观的破碎度和分离度与LST呈极显著正相关;LST对景观优势度的影响最敏感,建设用地优势度与LST呈极显著正相关,而林地、水体优势度增加则能有效降低地表温度.探索性空间数据分析(ESDA)及尺度效应分析表明:巢湖流域地表热环境具有显著的空间聚集性,且LST空间格局具有明显的尺度效应,巢湖流域地表热环境效应相关研究的适宜幅度域为2～4km.通过规划合理配置林地和水体景观类型资源,加强建设用地增长调控,优化景观格局是抑制流域地表热环境效应的主要对策.%Based on the Landsat ETM+/TIRS image data, the inversion of Land surface temperature (LST) were conducted by using atmospheric correction method, and the evolution of land surface thermal environment effect was analyzed from 2000 to 2013 in Chaohu Basin. Cell samples at five scales level were obtained through gridding method, the scale effect of LST and its correlations with the surface landscapes were further quantitatively analyzed at specified scales. The results showed that high LST area was mainly concentrated in Hefei city and spread to its peripheries, while low LST area shrank gradually due to turning into moderate LST area from 2000 to 2013 in Chaohu Basin. As a result, the land surface thermal environment effect was enhanced. The correlation analysis between LST and landscape structure indicated that the construction land landscape was the main reason for the land surface thermal environment effect in the basin, while water and woodland landscapes could restrain LST effectively. The influence of landscape pattern on land surface thermal environment effect was significant. The analysis of correlations indicated that strong significant negative correlations were observed between the fragmentation and separation indices of construction and rural land landscapes and LST, but there was significant positive correlation between the fragmentation and separation indices of woodland and water landscapes and LST. LST is most sensitive to the change of landscape dominance index. The dominance index of construction land was significantly positively correlated with LST, while increasing the dominance of woodland and water could significantly reduce the surface temperature. Analysis of the scale effect and Exploratory Spatial Data Analysis (ESDA) for LST indicated that the LST had a significant spatial clustering characteristic in Chaohu Basin, and that the spatial pattern of LST had an obvious scale effect. 2～4km was the suitable extent of the researches on land surface thermal environment in Chaohu Basin. Countermeasures derived from the study for restraining the surface thermal environment in Chaohu Basin included: allocating ecological functional landscape resources, such as woodland and water landscapes, reasonably through planning; strengthening regulations on the growth of construction land; and finally optimizing the landscape pattern.