中文摘要：

基于我国南方六省国家气象台站历史气象资料、1∶10万土地利用/覆被数据和NCEP再分析气温资料,通过比较气温变化在不同观测环境气象站之间的差异,分析中国南方三种主要土地利用/覆被类型对气温趋势的影响。结果显示：土地利用/覆被类型对气温趋势具有稳定的影响,建设用地的年均温、年均最高和最低气温的升温幅度均最高,耕地次之,林地最小。进一步利用再分析资料剔除区域大尺度气候背景影响后,建设用地的年均温升温趋势仍最大（0.105℃/10a）,其次是耕地（0.056℃/10a）,林地的升温趋势最小（-0.025℃/10a）,且为负。这表明对于研究区气温的升温趋势,林地具有抑制作用,建设用地具有增强作用,且增强作用较耕地强。林地的各季节平均气温的变化幅度同样低于非林地。

英文摘要：

In this article we have analyzed the impact of land cover types on climate warming in southern China based on observation data of 136 reference meteorological stations＇from China Meteorological Bureau in Southern China, the 1：100000 national land cover data from the 1980s to 2005 and NCEP/DOE AMIP-Ⅱ Reanalysis. We extracted the underlying surface of 3-km radius buffer zones around the meteorological stations in different historical periods, and distinguished the observational environment of the meteorological stations. Then, annual, seasonal and monthly changes of air temperature are analyzed. We compared the difference of temperature change at meteorological stations with different observational environments which are respectively cropland, built-up land and woodland, and drew some conclusions about the impact of land cover types on climate warming in southern China. The result shows that, in southern China, among the three main types of land cover in the study area, the changing trends of the annual average, annual average maxi- mum and the annual average mind.mum temperature in built-up areas are most significant, while those in cropland area are moderate, and those in forest area are least significant. By analysis of OMR values which are observed minus data of NCEP/DOE AMIP-Ⅱ Reanalysis, we found that the changing trend of annual average temperature in built-up land is still most significant （0. 105℃/10a）, followed by cropland area （0.056℃/10a）, and forest are- a （-0. 025℃/10a）. So, it is concluded that forest plays an inhibitory effect in climate war- ming, while built-up land and cropland play an enhanced effect, and urban areas enhance the climate warming more than cropland. We also conclude that the changing trends of average seasonal and monthly temperature in woodland are less significant than those of any other land cover types.