中文摘要：

研究区位于青海柴达木盆地及其邻近地区，本文利用MODIS的植被覆盖度栅格数据，对区内的286个表土花粉样品进行覆盖度估算的误差与精度检验，并且选取了尕海湖和青海湖6．0kaB．P．，2．5kaB．P．和现代的3个时间片段进行裸地、草地和乔木覆盖度变化的研究。冰消期以来的气候变化必然导致荒漠-草原-森林草原的边界发生位移，钻孔第四纪花粉数据的植被覆盖度重建是一个新的尝试，它可以为该地区植被带的位移和季风变化提供重要的环境替代指标。本文采用经验正交函数（EOF）对多变量花粉数据进行了组合变换，然后采用弦距离类比法获得覆盖度的定量重建值。两个钻孔的花粉-覆盖度转换结果清楚地揭示出柴达木盆地6．0kaB．P．时期较为湿润，而2．5kaB．P．和现代较为干旱，其中青海湖在6．0kaB．P．的乔木覆盖度达到6．13％，而到了2．5kaB．P．和现代降至0．5％左右。尕海的草地覆盖度在6．0kaB．P．达到46．87％，而到了2．5kaB．P．降至21．75％和现今的0％。本研究结果证明，采用表土花粉数据作为标准数据集来进行古代植被覆盖度转换是可行的，该方法相对于A／C比值对指示干旱程度方面更具有说服力，在季风边缘区和农牧交错带的环境变迁定量化研究方面具有较大的潜力。

英文摘要：

The studied area includes the Qaidam Basin, Qinghai Province and its neighboring regions. The research is focused on the accuracy test of pollen-based vegetation cover estimation including bare, herb and tree, for 286 surface pollen samples. The vegetation cover reconstruction is applied to the periods of 6.0kaB. P. ,2.5kaB. P. and 0kaB. P. from cores in Gahai and Qinghai lakes. The Qaidam area is characterized by transitions among desert, grassland and forest,which might have shifted since the Last Glacial Maximun resulted from climate fluctuation and monsoon changes. The present land-cover reconstruction is a new attempt and can be used as an environmental proxy for evaluating the climate changes and monsoon circulation. We use here the empirical orthogonal functions （EOF） to replace the multivariable pollen data, and then reconstruct quantitatively the past vegetation cover by means of the best analogue with the chord distance. The reconstruction shows that the climate was wetter at 6. 0kaB. P. In the Qinghai lake the tree cover was as high as 6. 13% at 6.0kaB. P.,and fell down to ca. 0. 5% at 2. 5kaB. P. and 0kaB. P. The Gahai lake result shows that the grass cover attained 46.87% at 6.0kaB. P., and decreased to 21.75% at 2.5kaB. P. and 0% at present-day. The present work testified that the pollen-based vegetation cover reconstruction is feasible as a land cover proxy. The vegetation cover reconstruction provides more abundant information and more reliable estimation on the dryness index in comparison with A/C ratio. The method has a high potential application on quantitative studies of climate changes in the summer monsoon zone and the farming-pastoral boundaries.