Institute of Geographic Sciences and Natural Resources, Chinese Academy of Sciences, Beijing 100101, China.
College of Resources Science and Technology, Beijing Normal University, Beijing 100875, China.
Sci Rep. 2017 Jan 30;7:41275. doi: 10.1038/srep41275.
Historical land use information is essential to understanding the impact of anthropogenic modification of land use/cover on the temporal dynamics of environmental and ecological issues. However, due to a lack of spatial explicitness, complete thematic details and the conversion types for historical land use changes, the majority of historical land use reconstructions do not sufficiently meet the requirements for an adequate model. Considering these shortcomings, we explored the possibility of constructing a spatially-explicit modeling framework (HLURM: Historical Land Use Reconstruction Model). Then a three-map comparison method was adopted to validate the projected reconstruction map. The reconstruction suggested that the HLURM model performed well in the spatial reconstruction of various land-use categories, and had a higher figure of merit (48.19%) than models used in other case studies. The largest land use/cover type in the study area was determined to be grassland, followed by arable land and wetland. Using the three-map comparison, we noticed that the major discrepancies in land use changes among the three maps were as a result of inconsistencies in the classification of land-use categories during the study period, rather than as a result of the simulation model.
历史土地利用信息对于了解人类土地利用/覆盖变化对环境和生态问题时间动态的影响至关重要。然而，由于历史土地利用变化的空间表达不明确、完整的主题细节和转换类型的缺乏，大多数历史土地利用重建都不能充分满足充分模型的要求。考虑到这些缺点，我们探讨了构建空间明确建模框架（HLURM：历史土地利用重建模型）的可能性。然后采用三图比较法验证预测重建图。重建表明，HLURM 模型在各种土地利用类别的空间重建中表现良好，其优值（48.19%）高于其他案例研究中使用的模型。研究区域最大的土地利用/覆盖类型被确定为草地，其次是耕地和湿地。通过三图比较，我们注意到三幅地图中土地利用变化的主要差异是由于研究期间土地利用类别的分类不一致，而不是由于模拟模型造成的。
