南极半岛地区地震勒夫波群速度与岩石圈结构

本文依据南极智利贝纳尔多奥伊金斯将军站的来自南桑德韦奇群岛的Ｍｓ＝６．４级地震面波资料，采用最新的适配滤波频时分析技术和网格频散反演方法，计算了南极半岛地区的勒夫波群速度频散，并获得了该区岩石圈结构模型。结果表明，西南极南极半岛地区地壳为成层结构，可分为三层，其分层厚度分别为５ｋｍ、８ｋｍ和１０ｋｍ。地壳厚度为２３ｋｍ，上地幔顶部速度为５．３２ｋｍ／ｓ。在深度５３ｋｍ以下，地幔介质中速度逐步降低为５．１１ｋｍ／ｓ和４．９ｋｍ／ｓ，最低值达４．８ｋｍ／ｓ，可以推断南极地幔亦为成层结构。     

改进SA方法的卫星视频图像超分辨率重建

为了提高卫星视频图像的质量,针对SA方法中无法实现小数倍超分重建和"黑色格网"问题,提出了改进的SA超分辨率重建方法。对于小数倍超分问题,首先建立参考帧低分辨率图像坐标系,以超分倍数构建参考帧高分格网,按表达的物方元素一致的原则,把所有低分辨率图像和高分辨率图像的坐标系统一在一个参考坐标系下;对于"黑色格网"改善问题,在使用运动估计参数时,为了减少由于运动补偿造成的误差,取消传统SA方法中四舍五入的运动参数补偿方法,改为以待求点为中心确定容许误差,通过容许误差选择参估点,用参估点计算高分辨率网格像素,以减少"黑色格网"现象。利用仿真和真实的卫星视频数据实验验证了该方法的有效性。     

多网融合的城市网格化治理解决方案

为加快构建源头治理,把城市内的人、地、事、物、组织等社会治理内容全部纳入网格,实现资源整合进网格、责任落实到网格、问题解决于网格,推动社会治理力量下沉,本文结合北京市怀柔区网格化社会治理的模式,提出融合基础数据、办理事项、办理流程、网格划分、热线电话、网格队伍、考评机制、指挥体系、信息系统的方案,以"多网合一、多元联动"为主要特点,有效地解决了网格化城市治理的诸多问题.     

参数优化随机森林算法的土地覆盖分类

针对随机森林算法进行土地覆盖分类时无法确定参数组合以得到最优分类结果的问题,该文提出了两种随机森林算法的参数优化方法。以北京市昌平区为研究区,应用Landsat TM影像,实现了基于光谱值、纹理特征和专题特征的随机森林土地覆盖分类。采用改进网格法和遗传算法对随机森林算法的参数进行选择与优化,比较了改进的网格法和遗传算法方法找到的参数组合最优解,并将优化参数后的随机森林算法与传统的最大似然法及未经参数优化的随机森林算法对比。实验结果验证了随机森林算法在土地覆盖分类上的适用性和稳定性,且该文提出的基于参数优化的随机森林算法能得到更高的分类精度。     

Implementation of a Conservative Two-step Shape-Preserving Advection Scheme on a Spherical Icosahedral Hexagonal Geodesic Grid

An Eulerian flux-form advection scheme, called the Two-step Shape-Preserving Advection Scheme(TSPAS), was generalized and implemented on a spherical icosahedral hexagonal grid(also referred to as a geodesic grid) to solve the transport equation. The C grid discretization was used for the spatial discretization. To implement TSPAS on an unstructured grid,the original finite-difference scheme was further generalized. The two-step integration utilizes a combination of two separate schemes(a low-order monotone scheme and a high-order scheme that typically cannot ensure monotonicity) to calculate the fluxes at the cell walls(one scheme corresponds to one cell wall). The choice between these two schemes for each edge depends on a pre-updated scalar value using slightly increased fluxes. After the determination of an appropriate scheme, the final integration at a target cell is achieved by summing the fluxes that are computed by the different schemes. The conservative and shape-preserving properties of the generalized scheme are demonstrated. Numerical experiments are conducted at several horizontal resolutions. TSPAS is compared with the Flux Corrected Transport(FCT) approach to demonstrate the differences between the two methods, and several transport tests are performed to examine the accuracy, efficiency and robustness of the two schemes.     

Exploiting Illusory Grid Lines for Object-Location Memory Performance in Urban Topographic Maps

In order to be successful in spatial orientation tasks, people need to recall locations and configurations of spatial objects from their memory. This understanding of geographic space often arises from experience with cartographic media representing topographic and topological information by graphic symbols. Learning spatial information from graphic media is influenced by different perception-based grouping effects distorting the accuracy of spatial object-positions and their relations. Such geometric inaccuracies can be softened by adding a grid layer, which regionalizes the map and can be used as an additional orientation pattern. This grid layer usually consists of solid lines and overlays semantic information. The present paper reports the results of two empirical studies on object-location memory (OLM) performance. In these studies, the amount of visual detail of the grid layer was reduced. By positioning the grid layer below specific urban topographic objects (study 1), the grid pattern was graphically interrupted. These interrupted grid lines were completed by cognitive completion mechanisms (illusory grid lines) described in the Gestalt principles of closure and continuation. The second experiment examined the maximum grid line gap that is closed by cognitive line completion and keeps an advantage for OLM (study 2).     

基于主成分变换模型的DEM格网聚合及其误差分析

利用主成分分析揭示变量之间关系的特性,进而提出一种既能保证较高精度又能较好地保持地形形态特征的DEM格网聚合方法。首先根据主成分变换模型推导DEM格网聚合数学公式,构建主成分聚合模型;然后以30m分辨率DEM转换为90m分辨率DEM为例,根据格网点属性间的权重关系聚合重构DEM。在此基础上,以均值聚合和双线性重采样聚合方法为比较对象,从聚合前后的检查点高程偏差的统计描述、空间分布与自相关性、地形形态保持程度方面分析3种聚合策略下重构DEM的误差特性。最后运用描述统计、半变异分析和等高线套合方法,定量评价主成分聚合重构DEM的质量效果。试验分析结果表明,同均值聚合和重采样聚合相比较,该方法重构的DEM既能保持较高精度,又能很好地保持地形形态特征。     

基于区间树的等高线提取算法研究(英文)

This paper proposes a new algorithm for determining the starting points of contour lines. The new algorithm is based on the interval tree. The result improves the algorithm＇s efficiency remarkably. Further, a new strategy is designed to constrain the direction of threading and the resulting contour bears more meaningful information.     

基于层次化P2P协议的网格空间数据库系统模型

针对传统空间数据库技术的不足,充分结合网格空间数据管理的新特点,提出了一种适合于网格环境的空间数据库系统模型(grid peer-spatial database management system,GPeer-SDBMS)。该模型运用P2P协议Tapestry构建了基于网格的层次化分布式空间数据库,不仅利用数据库的模式差异将网格空间数据库划分为由相同模式节点组成的多个虚拟节点集合,实现了空间数据的分布式存储,而且还对查询算法进行了有效的改进,并通过实验验证了查询算法的高效性。     

基于CTS的大地形仿真应用研究

大地形仿真是研究三维视景仿真的一个重要问题。针对三维地形仿真应用的现状,本文结合LOD技术、数据分块技术和数据管理技术,采用基于CTS软件的方案解决大地形仿真应用问题。该方案可以有效解决基于MultiGen软件的传统方案中的纹理和地形边界不连续以及内存容量限制等问题。通过实例验证了该方案切实可行。