Font Size: a A A

Spatio-Temporal Data Modeling And Visualization For Dynamic Phenomenon

Posted on:2010-03-29Degree:DoctorType:Dissertation
Country:ChinaCandidate:Z W HeFull Text:PDF
GTID:1220330332985671Subject:Cartography and Geographic Information System
Abstract/Summary:PDF Full Text Request
The object of study in this dissertation is dynamic phenomenon, that is a great variety of geographical or natural phenomena, which occur in the space closely related to the human living environment, such as dust storms, clouds, fog, haze, smoke, diffusion of polluted gas, tornadoes, the ocean currents, and so on. Dynamic phenomenon is a dynamic, varied, open and complex system with ambiguous borders. It is different from the usual moving objects which have the clear boundaries and move regularly. Dynamic phenomenon is closely related and affects the human economic, political, military and social activities.With the development of geo-information science, it has been applied in many fields, such as disaster prevention and mitigation, environment protection, military applications, resource management, engineering construction, transportation navigation, etc. Currently, geographical information system(GIS) focus on static surface features and moving objects, however, they lack the dynamic performance and continuous evolution and reasoning mechanism. Because dynamic phenomenon is changeable, the current GIS can not express it completely, that is, they can only describe the two-dimension and static objects, thus lost a lot of information. Usually, temporal GIS add temporal dimension on to spatial dimensions, take time dimension as the benchmark of spatial object rigid movement. And that is not suitable for the description of continuous changing process of dynamic phenomena.In order to organize and manage spatio-temporal data of dynamic phenomena, it is necessary to create a spatio-temporal data model for dynamic phenomena, which is the core of the spatio-temporal database. Moreover, it is the foundation of the description of dynamic phenomena. The current spatio-temporal data model can not be effective in organizing and managing the spatio-temporal data which is created in the dynamic process. The purpose of this study is to establish a mechanism that can describe the continuous changing process of objects and phenomena, to organize and manage the spatio-temporal data, to realize the multi-dimension dynamic visualization in GIS and use them in application system.In this dissertation, the four aspects listed as follows are discussed:1) In order to build spatio-temporal data model for multidimensional dynamic phenomena, a describing mechanism for complex dynamic phenomena process is introduced into geographic information science, which embeds behavior model of continuous dynamic changing object into existing spatial information representation system. Based on spatial cognition, the mechanism for dynamic phenomena description and representation are discussed and BM-STDM spatio-temporal data conceptual Model, Logical model and implementation framework are constructed from the point of view of spatio-temporal semantics. The behavior or changes of spatial object and phenomena be brought into spatial recognition system. From the view of spatio-temporal recognition, some concepts, such as event, state, continuous event, transient event and event point, are defined and 4D spatio-temporal reference system is described then. Behavior model is used to describe the. development process of dynamic continuous changing object.2) The method for managing and organizing the spatio-temporal data of complex and dynamic object is proposed. Behavior model is a high level summary of the changing process of spatial object, which describes spatio-temporal behavior characteristics of dynamic phenomena. This dissertation describes three behavior models for dynamic phenomena. In this dissertation, polluted gases diffusion in urban is studied as example and the gas diffusion behavior model is built. Meanwhile, the model building process and methods are also introduced. Based on behavior model and spatio-temporal data model, industry application models can be embedded into GIS application system. By referring to the management of model library, access and transfer technology for spatio-temporal data model and professional application model library is built. This dissertation put forward using an octree method for the organization and storage of spatio-temporal data, and discusses operations and query analysis for spatio-temporal data, such as create, modify, update and destroy.3) A new 4D spatio-temporal visualizing technology for dynamic field data is put forward. In this dissertation, the LOD model of multi-scale representation for complex phenomena are built to satisfy the requirement of different application and visualizing condition. This model can also simulate and represent complex dynamic phenomena in 3D virtual earth scene. The experimental data is obtained from simulation of physical model. Applying new technology in computer graphics, a dynamic partition projection rendering method for spatio-temporal data is proposed in the research of real-time 4D rendering technology of dynamic phenomena field data. Some other multi-dimensional visualization methods, such as ray cast method and GPU based 3D texture mapping method, are compared with this method in experiment.4) Solve the integration issues of complex dynamic object and specific application system. In order to build experimental platform, foregoing spatio-temporal data model and multi-dimensional real-time rendering technology are integrated into virtual earth platform environment of GeoGlobe. The embedding mechanism for multi-dimensional representation methods of dynamic object into application system is studied, which can be used to manage and schedule dynamic object data and behavior model. In this dissertation, the diffusion model of urban polluted gases is studied to test and validate the proposed theories and methods.This dissertation makes a preliminary research in the modeling and expression of spatio-temporal data in GIS, and does some meaningful explorations in the expression of the continuous dynamic phenomena. Experimental results show that the proposed BM-STDM spatio-temporal data model can effectively organize and manage the spatio-temporal data of the polluted gas diffusion and can reduce the amount of storage space. Through the behavioral model, it becomes possible to describe the continuous dynamic phenomena and to embed a variety of business application models into GIS. Experimental results also show that the proposed dynamic partitioning projection rendering method can better visualize dynamic phenomena in the virtual earth scene and it can satisfy the need of real-time visualization. Research in this dissertation can be achieved in specific applications to solve a real-time, dynamic, three-dimensional mapping and scheduling problem, can visualize the scalar data and vector data dynamically, to achieve a seamless integration and multi-scale representation of the dynamic phenomena with virtual earth scene.The main Creative points in this article are the following:1) A model describing the continuous dynamic phenomena is proposed. On the exploration of organization and management of spatio-temporal data, the BM-STDM is presented, which can realize modeling and expression of dynamic phenomena and can organize and manage the massive spatio-temporal data of dynamic phenomena. Moreover, it has a relatively small amount of storage space and can do access and query analysis to the spatio-temporal data quickly.2) A mechanism and framework, which can embed the industrial application model seamlessly into GIS, is proposed. The behavioral model can realize the more completely combination of industry application models and GIS in Semantic aspect. And this dissertation described the specific methodology and framework for the combination.3) This dissertation establishs LOD models for complex dynamic phenomena, and realizes the expression of multi-scale dynamic phenomena, which can meet the needs of different applications in the limited visualization capabilities environment for computing, storage and rendering.4) A 4D spatio-temporal visualization technology for dynamic phenomena is proposed, and the multi-scale expression and real-time visualization of complex dynamic phenomena in virtual environment is realized. The mechanism can embed multi-dimensional performance methods of dynamic phenomena into application system. This dissertation also solves several problems in the process of embedding dynamic phenomenon into virtual earth platform, such as seamless integration, real-time dynamic rendering and scheduling.The development trend of geographic information technology is in the direction of network and multi-dimension development. The research of dynamic phenomenon will be a new peak in the study of multi-dimensional GIS. This dissertation makes a useful try for the expression of dynamic phenomena, spatio-temporal data organization, management and visualization. The development of geographic information technologies will promote the future Digital Earth be more dynamic and sophisticated.
Keywords/Search Tags:spatio-temporal data model, dynamic phenomenon, continuous changing process modeling, virtual earth scene, real-time rendering, spatio-temporal data organization and management, behavioral model, multi-scale representation model
PDF Full Text Request
Related items