中文摘要：

对黄土沟谷形态结构和发育演化起核心控制与标定作用的,是一系列具有自组织结构的特征点集。针对该特征点集,提出黄土沟谷特征点簇的概念,对其组织要素和基本特征进行了分析;研究创建一种采用非结构化、且能够有效表达拓扑结构与空间信息的黄土沟谷特征点簇空间结构模型;对基于DEM数据获取的一个典型黄土小流域沟谷特征点,运用创建的点簇模型建立了具体的黄土沟谷特征点簇;最后,以点簇中的特征点追溯查找为例,验证了所创建的黄土沟谷特征点簇的应用价值。

英文摘要：

In the loess gullies, a series of key features points set play important roles such as central control and calibration for its Morphology and development. For the gully feature points set, systematically studying the scientific meaning, building the structural integrated data model, and studying the loess gully landform＇s spatial differentiation of evolutionary tracks and development by using the data model have important scientific significance and practical application value. So, the concept meaning of loess gully feature points cluster is presented and analyzed. Loess gully feature points cluster is a points＇ organic whole which plays a decisive role（such as gully head point, break point, et al.） or Calibration role of the terrain（such as gully node point, runoff source point, et al.） in a loess watershed. Based on DEM data, the gully feature points of a typical completed watershed in the loess landform are extracted. Using object-oriented modeling ideas, an unstructured file storage mode of the loess gully points cluster is researched and created. The data model can effectively describe a performance points cluster topology and spatial information. The loess gully points cluster model of test area is achieved through programming. A case study as the feature points retrospective sought, the processing efficiency is higher by comparing.shp data mode. The research results show that： the loess gully feature points cluster retains the key character of loess gully, and especially has obvious advantage on expressing the morphological structure characters. The points cluster space model uses the tree and non storage structure, so its structure is clear and its description information storage is simple. At the same time, the points cluster model application has higher efficiency in the valley spatial structure analysis.