多源多尺度DEM地表形态精度量化分析研究

The existing researches of DEM quality analysis mainly focuses on the accuracy of elevation value. Nowadays, DEMs are constructed from multi sources such as contour digitalization, Lidar based altimetry and so on. As their multi scale and multi sources characteristics, there are less effective technologies which could analyze and assess the morphology accuracy of DEMs. It may become the technology bottleneck of the DEM uncertainty evaluation. In this research, based on the comprehensive application of qualitative and quantitative methods, macroscopic and microscopic methods, analytical and comprehensive methods, Shaanxi province with various kinds of landforms are taken for example. With the perspective of high-low sequence of elevation, convex and concave characteristics of local slope, spatial relationship of local terrain, structural characteristics of macro-relief, the research containsthe following parts: studying on the concept model and accuracy index of DEM surface morphological errors, analyzing the formation mechanism of DEM surface morphological errors, constructing the information map type evaluation criterion of DEM surface morphological accuracy with the characteristics of terrain information cognition, establishing the quantitative method of DEM surface morphological errors based on space field model, designing the prototype system for evaluating DEM surface morphological accuracy , mainly solving the problem of the evaluation for surface morphological accuracy and usability from multi-source and multi-scale DEMs.This research will further prefect the theory and method of authentic assessment of DEM terrain representation and the construction and application ability of DEM with scaling sequence will be improved significantly.

地表形态是派生一切地形信息的本源。现有的DEM质量分析主要关注于DEM的高程数值精度，面对当前丰富的多种来源、多种尺度DEM数据，缺乏有效技术分析和评价其地表形态精度，已成为推演DEM应用不确定性的瓶颈问题。本课题将综合运用定性与定量、宏观与微观、解析与综合相结合的研究方法，以地貌类型多样的陕西省为例，从高程高低序列关系、局地坡面凸凹特征、区域地形空间关系、宏观地形结构特征等视角，研究DEM地表形态误差的概念模型与描述指标，剖析DEM地表形态误差的形成机理，构建具有地形信息认知特色的信息图谱型序列尺度DEM地表形态精度评价标准，建立基于空间场模型的DEM地表形态精度量化分析方法，研发DEM地表形态精度评价原型系统，重点解决多源多尺度DEM地表形态精度评价与可用性分析问题。本研究将进一步完善DEM地形描述真实性评价的理论与方法，有力提升序列尺度DEM的建设能力与应用水平。

现有DEM质量分析主要关注于DEM高程数值精度，面对当前丰富的多种来源、多种尺度DEM数据，缺乏有效技术分析和评价其表达的地表形态精度，已成为推演DEM应用不确定性的瓶颈问题。本研究将DEM表达地表形态的区域性失真原因归纳为格网DEM数据结构自身局限、地形信息采集不完整、DEM高程内插计算不严密等3类，在此基础上提出一种描述DEM地表形态误差的概念模型；从DEM地表形态误差的主要形成机理与主导因素展开研究，提出一种基于微分几何的DEM地表形态精度量化模型，一种基于地形描述误差的量化分析模型，一种基于场理论的DEM地形描述误差场模型；围绕DEM地形信息的尺度演化规律开展探索性研究；设计并完成了面向DEM地表形态精度评价的原型系统及序列尺度DEM构建与分析系统。项目组基于本研究取得的进展，继续进行拓展研究，在新型DEM模型设计方面提出一种基于SVO模型的新型DEM构建与应用技术，在面向DEM场模型精度分析方法方面提出一种基于改进归一化距离匹配函数的纹理分析模型。项目组共发表研究论文13篇（其中SCI/EI索引论文6篇），研究成果获中国地理信息科技进步二等奖1项。