中文摘要：

针对当前地貌分类较少关注微观地貌特征信息的问题,该文依据1∶25万DEM数据开展湖南省地貌形态特征分类的实验研究。首先,确定分类实验的最佳分析窗口为0.28km~2。其次,根据可以反映宏观形态特征的起伏度和切割度指数,可以反映微观形态特征的地形位置指数（TPI）,构建了分类指标体系并将湖南省按照地貌形态特征分为12种地貌类型。第三,地貌分形维数值是检验地貌形态特征的重要指标,选用研究区5个小流域并分别计算其分维数值,据此进行分类结果检验。研究结果说明该文的实验方法可以较好地体现微观地貌形态信息。

英文摘要：

Using DEM data with 1：50 000 as research materials, this paper focuses on the landform taxonomic features of Hunan Province of China. Firstly, this paper decides on the optimum analysis window with 21 ＊ 21 sizes, which is about equal to 0. 28 kmz in study area. Secondly, this paper employs three indices as classifying metrics, such as relief, surface incision and TPI, respectively. With these indices, the paper forms a taxonomic framework for landforms in study area. According to experiment, this paper gets 12 basic types for study area. Lastly, this paper uses 5 little watershed for a result test by using their fractal values. Although there are some potential deficits in our methods, this paper still indicates that the taxonomic framework can play a significant role in landforms partition.