QQ客服 官方微博 反馈留言

题名	基于CA和GIS的流域侵蚀产沙系统模拟研究
作者	原立峰
学位类别	博士
答辩日期	2006
授予单位	中国科学院水利部成都山地灾害与环境研究所
授予地点	成都
导师	周万村 ; 周成虎
关键词	复杂性 元胞自动机 地理信息系统 侵蚀产沙模型 流域模拟
学位专业	自然地理学
中文摘要	目前，国际上在复杂系统理论的背景下，利用元胞自动机模型模拟地理时空过程的复杂行为是地理建模的前沿地带。流域侵蚀产沙系统是一个典型非线性复杂动力系统，具有开放性和动态性、非平衡性和时空变异性、多尺度性、过程不可逆性和多重耦合性，以及非线性动力演化等特征。元胞自动机模型是一种时间、空间和状态都离散、空间上的相互作用和时间上的因果关系皆局部的网格动力学模型，是研究复杂系统的重要方法之一。将以上两者结合具有天然的合理性，元胞自动机的时空动态特征及强大计算能力能够为流域侵蚀产沙复杂系统模拟提供一种新的思路。论文以流域侵蚀产沙系统的复杂性为理论依据，以元胞自动机（Cellular Automata）、地理信息系统（GIS）和遥感等为技术手段，对不同时空尺度上的侵蚀产沙现象进行了动态模拟。在坡面尺度上，利用元胞自动机方法，同时结合自组织理论，对细沟的形成过程进行了时空动态模拟，生动形象地再现了坡面细沟侵蚀的发生、发展和演化过程；在流域尺度上，构建了基于次暴雨的流域侵蚀产沙元胞自动机模型——CASEM，并利用Visual Basic 6.0编程语言和ArcObject GIS组件库，采用面向对象的程序设计思想，开发了用于流域模拟的计算机软件系统；以陕北延安地区燕沟流域作为实证研究区域，利用CASEM模型对该流域1999年7月10日的一次降雨侵蚀产沙过程进行了动态模拟。本文的创新点在于将元胞自动机的理论与方法和土壤侵蚀学科相结合，动态的模拟和预测侵蚀产沙系统的发展演化过程，为侵蚀产沙研究开拓新的研究手段，并为侵蚀预报模型的定量化研究拓宽途径，在方法上具有一定的创新性。另外，论文利用综合的观点和学科交叉的方式对流域侵蚀产沙系统进行了深入地探讨，丰富了流域侵蚀理论研究。论文取得的主要研究成果如下：（1）关于侵蚀产沙模型的分类。从复杂性科学的理论观点出发，系统地论述了流域侵蚀产沙的复杂性本质特征；在此基础上，将流域侵蚀产沙模型进行了重新分类，分为经验模型、概念模型、物理模型（集总式）、分布式模型和复杂性模型等类型，提出了侵蚀产沙复杂性模型的概念，认为利用复杂性理论和方法所建立的模型都属于复杂性模型，模型的特点是能够反映侵蚀产沙系统的复杂性特征，研究方法先进，模型求解效率高。这种分类方法直观的反映了侵蚀产沙的建模历程，体现了人们对侵蚀产沙过程理解的深入和建模思想的成熟。研究认为，利用复杂性科学的理论和方法，运用现代GIS技术，并结合分布式模型建模思想建立的复杂性模型将是今后一段时间侵蚀产沙模型发展的新方向。（2）坡面细沟自组织形成过程。降雨作用下，坡面细沟侵蚀的发育过程实际上可以看作是一个将无序输入转化为有序输出的自组织过程；其中，降雨和微地貌对应系统的无序输入，而细沟沟网的形成则对应系统的有序输出；基于这种自组织理论思想，利用元胞自动机方法构建了坡面细沟侵蚀时空动态模型——CA_Rill，模型着重于微观机理的考虑，通过元胞间简单的局部相互作用关系就可以反映出坡面细沟从无到有的沟网形成过程，而不需要进行复杂的数值求解；通过模型敏感性分析认为，2厘米大小元胞能够较好的展现坡面细沟沟网的发育过程，在大雨强和较大地表糙率的条件下，更易于形成等间距平行排列的细沟。（3）流域侵蚀产沙系统元胞自动机模型CASEM的构建。研究构建了流域尺度的、基于次暴雨的侵蚀产沙CA模型——CASEM。模型对标准的元胞自动机进行了多方面的扩展，更多的考虑了流域侵蚀产沙系统内在的物理机制，以分布式模型的建模思想为指导，以元胞自动机为核心，将流域按照规则格网元胞进行空间离散化，每个元胞都以一组参数来表示不同元胞特殊的地表下垫面状况，并以此作为流域侵蚀产沙的基本计算单元。模型考虑分离和输移两种物理过程，每个元胞上的土壤侵蚀和泥沙输移遵循Meyer和Wischmeier（1969）提出的理论，即土壤侵蚀的输出量等于输入量与降雨—径流产生的泥沙量之和，且输出总量不得超过降雨—径流的输移能力。流域内径流分布、侵蚀与泥沙输移过程通过元胞及其邻居间的水、沙交换来实现，即在每一个元胞步长内，通过扣除每个元胞上的侵蚀量或加上沉积量不断更新元胞空间，从而实现流域侵蚀过程时空动态模拟。（4）CASEM模型软件系统的设计与开发。采用Visual Basic 6.0编程语言和ArcObject GIS组件库构建CASEM模型的计算机软件系统。将GIS和侵蚀产沙CA模型进行了紧密集成，该系统可以对指定流域的、次暴雨侵蚀产沙过程做出动态模拟和预测。系统功能包括图形显示、参数设置、流域模拟、统计分析和显示输出等部分。该系统具有界面友好，操作简单，对硬件配置要求不高，图形显示效果好等优点。（5）燕沟流域侵蚀产沙系统的时空模拟。选择延安燕沟赵庄小流域作为大比例尺研究对象，对该流域1999年7月10日的一次暴雨过程进行了模拟。模拟结果表明，CASEM模型结构简单，求解效率高、图形效果好、可以直观地展现次降雨情况下流域侵蚀产沙时空分布，充分体现了元胞自动机方法和分布式建模思想结合所展现的强大优势。实践证明该系统具有一定的实用性。
英文摘要	Nowadays, it has been taken as an international scientific frontier that using cellular automata at the background of the complex systematic theory model various complex behaviors of geographical spatial-temporal processes. Erosion and sedimentation system in catchment is a typical non-linear complex dynamic system, taken many complicated features, such as opening and dynamicity, non-equilibrium and space-time variability, multi-scales, process irreversibility and multi-coupling and non-linear dynamic evolvement, etc. Cellular automata is a grid dynamic model whose time, space and state all are discrete and whose cause and effect in time scale is local, which is one of the important research methods of the complex system. There is naturally reason for the integration between cellular automata and erosion system. Spatial-temporal dynamic feature and strong computing power of cellular automata can provide a new thought for erosion and sedimentation complex system simulation. This paper based on the complexity theory of erosion and sedimentation system in watershed, and dynamically modeled erosion phenomena on different spatial-temporal scale by way of various techniques, such as cellular automata, GIS and remote sensing, etc. On hillslope spatial scale, rill-forming processes were simulated by use of cellular automata model and self-organizing theory, and its development processes was vividly reproduced. On watershed spatial scale, CASEM, a watershed erosion system cellular automata model, was developed successfully. The software system of CASEM was designed and developed basing on Visual Basic programming and ArcObject GIS module library, all of this used the thought of oriented object programming design (OOPD). CASEM was applied to model once rainfall-runoff process on July 10, 1999 at YanGou watershed in Northern Shaanxi of P.R. China. Innovation point of this paper is that some correlative theory and methods of cellular automata and soil erosion discipline are firstly combined, and dynamically imitation and predict the evolvement of soil erosion is analysis, and also a new tool for soil erosion study and extend approach for soil erosion quantifying research is supplied. From the point of research method, the way used in this paper is fresh. Additionally, in this paper integration learning view and manner of discipline intersect research into soil erosion system in watershed is adopted, which enrich subject content and improve discipline development. So, in a sense, this paper has scientific innovation value at some extent. Main research fruits of this paper are as follows: III 中国科学院博士学位论文 原立峰 基于 CA 和 GIS 的流域侵蚀产沙系统模拟研究 (1) Erosion and sedimentation model classification. From the view of complexity science, complexity essence feature in watershed erosion and sedimentation system was systematically discussed. On the basis of these works, erosion models were reclassified into empirically-based model, conceptual model, physically-based model, distributed model and complex model. The concept of complex model was put forward. It includes all the models created by use of theory and methods of complexity science. The main characters of these models are that they reflect complexity feature of erosion and sedimentation system, which would propel and put ahead current research method due to high efficiency. The classification method can straightly show the modeling course of soil erosion, so that researchers can understand deeply erosion processes and mature modeling thought. The results indicated that complex model using theory and method of complexity science and combined modeling thought of distributed model is a new development direction in the future. (2) Self-organizing process of hillslope rill-forming. In fact, the growth processes of rill on hillslope under the action of precipitation can be taken as a self-organizing process, in which disorderly input such as rainfall and micro-landform was transmitted into an orderly output like the form of rill network. Based on these self-organizing theory thought, CA_Rill, a rill erosion spatial-temporal dynamic model, was developed by way of cellular automata model. It emphasizes micro-physical progress of soil erosion, and reflects the evolvement progress of rill on hillslope. Moreover, there is no need to solve complex mathematic formula for CA_Rill model. By model sensitivity analysis, this paper indicates that 2cm cell size can show finely rill-forming progress. The greater rainfall intensity and land roughness, the easier rill comes into being. (3) Development of watershed erosion and sedimentation cellular automata model. A cellular automata model based on watershed scale and once rainfall was developed, which named as CASEM. This model extended standard cellular automata model in many ways, and emphasize on inner physical mechanism in watershed erosion and sedimentation system. CASEM takes distributed modeling thought for guidance, and takes cellular automata for core, and discrete watershed space in regular grid. Each cell represents different soil surface state in a set of parameters, which have been taken as basis compute unit for erosion and sedimentation. CASEM considers two kinds of physical action forces: rainfall and runoff. Soil erosion and sedimentation transportation in each cell follows the theory put forward by Meyer and Wischmeier in 1969, i.e., output amount of soil erosion is equal to input amount of sedimentation produced by rainfall and runoff, and all output amount is less than transportation capacity of rainfall and runoff. Runoff distribution and transportation process of erosion and sedimentation represented through exchange of soil and sand among cell and its neighbor, that is, cell space is updated continuously by subtract erosion amount and add sedimentation amount in each cell step. (4) Design and development of CASEM model software system. Adopting Visual Basic 6.0 programming and ArcObject GIS module library, we developed compute software system of CASEM model. GIS and CA model were tightly integrated at the same time. This presented system can evidently model and predict soil erosion progress on the effect of one precipitation procession. System function includes graphical display, parameters set, watershed simulation, statistical analysis and display output, etc. System has several excellent features, such as a friendly user interface, simply operated and fine graphical effect, and so on. (5) Spatial-temporal processes simulation of erosion and sedimentation in YanGou Watershed. Zhaozhuang catchment was selected as a research object. Once rainfall-runoff process on July 10 in1999 was simulated. The results indicated that CASEM model have some advantages as follows: simple configuration, high efficient algorithm, fine graphical display and good model precision. CASEM can straightly display spatial-temporal distribution of erosion and sedimentation on the action of once rainfall process, and show strong advantage combining cellular automata method with distributed modeling thought. The general results show CASEM is useful.
语种	中文
学科主题	摄影测量与遥感技术
公开日期	2010-10-22
分类号	TV8;TP3
内容类型	学位论文
源URL	[http://ir.imde.ac.cn/handle/131551/2290]
专题	成都山地灾害与环境研究所_成都山地所知识仓储（2009年以前）
推荐引用方式 GB/T 7714	原立峰. 基于CA和GIS的流域侵蚀产沙系统模拟研究[D]. 成都. 中国科学院水利部成都山地灾害与环境研究所. 2006.

个性服务

查看访问统计

相关权益政策

暂无数据

收藏/分享

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。