基于软硬分类和CA-MAS的杭州湾滨海湿地演变模拟与优化研究

项目名称： 基于软硬分类和CA-MAS的杭州湾滨海湿地演变模拟与优化研究

项目编号： No.41201458

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 地理学

项目作者： 胡潭高

作者单位： 杭州师范大学

项目金额： 25万元

中文摘要： 杭州湾滨海湿地是我国重要的河口型滨海湿地，是环杭州湾地区重要的区域生态安全屏障，对环杭州湾地区的资源、环境、经济的可持续发展具有十分重要的意义。随着自然灾害的逐年增多以及社会经济的迅速发展，杭州湾湿地资源正面临着自然和社会经济发展的双重压力，因此，研究在自然和人为因素双重影响下的湿地变化趋势，对合理保护湿地资源，促进社会经济的可持续发展具有重要作用。本研究以地物类型丰富、受城市化影响严重、重要的湿地保护区- - -杭州湾滨海湿地为研究对象，针对地物类型繁多、类型间转换频繁等特征，利用多源遥感数据、基础地理数据和社会经济统计数据等，在遥感信息智能提取技术，元胞自动机和多智能体系统理论的支持下，研究建立杭州湾滨海湿地演变模拟优化模型，通过该模型模拟杭州湾滨海湿地未来近20年的演变情景，并在此基础上分析湿地演变机制和保护对策，为科学合理的保护和利用杭州湾湿地资源提供决策服务。

中文关键词： 湿地监测；多光谱遥感；软硬分类；景观格局；驱动机制

英文摘要： The Hangzhou Bay coastal wetland is an important estuarine-coastal wetlands, which plays a significant role on the sustainable development of resources, environment and economic in China Yangtze River Delta. In recent years, the natural disaster increased gradually and the economic developed fast, the Hangzhou Bay wetland resources are facing the dual pressures from natural and socio-economic development. Therefore, the research on the changing trends of wetlands under the double impact of natural and anthropogenic factors plays an important role for the reasonable protection of wetland resources and sustainable development of social economic. In this research, we choose Hangzhou Bay coastal wetland as study area which contains rich feature types and highly urbanized. Firstly, we analyze the characteristics of different types of wetlands, collect multi-source of remote sensing data, basic geographic data and socio-economic statistical data. And then based on the above preparation works, we build wetlands evolution simulation and optimization model in Hangzhou Bay under the support of remote sensing intelligent classification technology, cellular automata and multi-agent systems theories. Through this optimization model, we simulate the scene of Hangzhou Bay coastal wetlands nearly after 20 years. Finally, with t

英文关键词： Wetland Monitoring；Multispectral Remote Sensing；Soft and Hard Classifications；Landscape Pattern；Driving Factors