中文摘要：

历史时期土地利用／覆盖变化一直是国际关注的热点问题之一。研究历史时期土地利用／覆盖变化，不但可以认识和理解陆地表层景观环境在过去的变化过程，而且对当前的土地利用方式或预测未来土地利用／覆盖变化和发展都有重要的指导意义。依据使用的主要数据源、处理方法和研究思路，从历史文档资料、古地图、考察／调查报告、模型模拟与综合多源信息和多学科知识的方法等5个方面综述国内外历史时期土地利用变化研究方法的现状。根据不同数据源的特性，探讨分析利用不同方法开展历史时期土地利用变化研究的优缺点及其不确定性：①历史文档资料种类繁多，包含的信息量巨大，但存在很多噪音和遗漏，且缺乏直观的空间信息；②古地图可以提供比较直观的空间位置信息，但是历史时期的绘图技术落后，且土地利用类别划分较粗；⑧过去的野外考察／调查资料记录比较规范，但时空尺度较小，难以满足较早时期的研究需要；④利用模型进行模拟研究可选择的时空尺度范围较大，但模型的选择和建立比较复杂，对驱动数据和初始值要求较高。最后，提出只有综合多源信息和多学科的知识才能比较完整地理解历史时期土地利用变化过程，融合多源信息和多学科知识的方法将是今后进行历史时期土地利用变化研究的主要思路。

英文摘要：

Historical land use and cover change have been one of the hot topics of international concerns. Re- construction and research of historical land use and cover change can understand the process of past changes in landscape environments of land surface, and have an important guiding role in the current land use patterns and in predicting the future land use and cover change. Based on the main data sources used and data process metheds and study ideas, the research status of historical land use change are reviewed both at home and abroad in terms of historical documents, historical maps, inspection and investigation reports, model simulation and integrated multi- source data and multidisciplinary approaches. Then, the advantages and disadvantages and uncertainties of the vari- ous methods used to carry out the research of historical land use change are discussed and analyzed according to the characteristics of different data sources and methods： （1）We have wide range of historical documents and huge quan- tity of information, but there are a lot of noise and omissions, and lack of visual spatial information; （2）Historical maps can provide a relatively intuitive spatial location information, but the mapping technology was poor and the land use categories were coarse in historical period; （3）Past inspection and investigation records with a smaller spa- tial and temporal scales are fairly standard, but they are difficult to meet the needs of the earlier period; （4）Using the models to implement the simulation in different spatial and temporal scales is relatively easy, but more input da- ta are needed. Finally, the integrated multisource data and multidisciplinary approach are considered as the only way to understand the process of historical land use change essentially and completely, and it will be the main ideas to study the historical land use change in the future.