The authors use a web crawler to retrieve all periodical articles from CNKI between the 1950 s and 2016 and then parse the abstracts of 293368 articles about grassland deterioration by word segmentation, location matching and other methods. The authors also construct a research hot regions extraction model of grassland deterioration in China based on a comprehensive research hot regions index of toponyms and then analyze the spatial pattern and dynamic change in research hot regions of grassland deterioration in China. The research shows the following：（1） The spatial heterogeneity of grassland deterioration in China can be effectively described by a model of grassland deterioration based on the comprehensive research hot regions index.（2） The research hot regions of grassland deterioration are mainly distributed in most regions of Inner Mongolia, Xinjiang, Qinghai, Tibet, Gansu and other provinces. The northeastern region of Inner Mongolia（such as Hulunbeier） and the eastern region of Inner Mongolia（such as Xilin Gol, Chifeng and Wulanchabu） are significant hot regions in the study of grassland deterioration.（3） The number of high research hot regions increases from 81 in the 1950 s to 99 in the 2000s; the area increases from 1.038 million km2 to 1.146 million km2. The degree of hot for grassland deterioration research in 197 counties showed an upward trend. This paper also discusses the relationship between the region of research hot regions and the region of grassland deterioration and then indicates the differences between them in time matching, space matching and concept matching.
中国是世界上水土流失最为严重的国家之一。准确掌握既有水土流失研究的空间分布格局是一项基础性工作。以中国知网学术期刊数据库作为数据源,应用自然语言处理方法,对1980-2017年中国水土流失研究地区进行了地名信息提取及研究热点建模;继而应用RUSLE模型模拟,得到全国土壤侵蚀强度的空间分布;在上述研究基础上,对研究热点地区与侵蚀强度之间的空间耦合关系进行了对比分析。结果表明:(1)1980年以来,中国水土流失研究热点区主要分布在黄土高原及贵州高原,涉及陕西、宁夏、内蒙古、甘肃、贵州以及黑龙江等省区;中等及以上热度的县(区、市)共171个,占全国国土总面积的5.33%。(2)RUSLE模型模拟表明,严重的土壤侵蚀主要分布在黄土高原及云贵高原,涉及陕西、宁夏、甘肃、山西、贵州、云南、四川等省区;侵蚀模数大于20 t hm-2 a-1的县(区、市)共251个,占全国国土总面积的7.04%。(3)研究热点地图与水土流失强度模型模拟地图之间存在空间差异。对特定空间耦合模式的分析有助于判断科研资源配置的合理性。
