中国湿地分布的时空变化模拟研究

项目名称： 中国湿地分布的时空变化模拟研究

项目编号： No.41271423

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 天文学、地球科学

项目作者： 牛振国

作者单位： 中国科学院遥感与数字地球研究所

项目金额： 70万元

中文摘要： 湿地分布及其时空变化，对于全球变化、生物多样性、环境与健康等都具有重要意义。影响中国湿地分布及时空变化的因素复杂、且随时空不同而对湿地具有不同的影响程度；项目拟利用Random forest方法对这些因素进行定量辨析，给出这些因素对湿地影响的贡献度；并基于已有研究成果，利用SOFM方法进行中国湿地变化分区的研究；同时从湿地发生学角度，研究构建中国湿地的潜在分布模型；在此基础上，以现有中国湿地数据（1978-2008年）为基础，分别基于CA模型和logistic模型研究建立中国湿地分布的时空变化模拟模型，模拟中国湿地空间格局的演变；基于多源遥感数据分类和相关调查等辅助地学数据，对模拟模型进行验证，并从运行效率和预测精度等方面对模型进行对比分析。模型方法不仅将为全国大范围湿地制图探索新的途径，也为满足中国湿地管理的需要提供科学支持。

中文关键词： 湿地模拟；中国湿地变化；湿地遥感；时间序列；驱动力分析

英文摘要： Wetlands distribution and its spatio-temporal changes are extremely important to global changes, biodiversity, environment and health. The impact factors those influence the spatial and temporal distribution of China wetlands are much more complicated, which have various impacts on wetland changes along with space and time. This project is designed to study the model that was employed to simulate and predict China wetlands spatial distribution. So the project plans are as follows: firstly, the random forest method will be employed to quantitatively differentiate and analyze the aforesaid impact factors to determine their contribution to China wetland changes. Secondly, based on the previous researches, the project will also plan to continue the research of regionalization of china wetland changes by SOFM method. Thirdly, wetlands potential occurrence model was redesigned by reference of related researches at home and abroad, and then used to predict wetlands potential distribution across China, which was validated according to existed wetland map in 1978 and other maps. After that, China wetland changes model, which is based on CA model and logistic model, was studied elaboratively and employed to predict the China wetland spatial distribution from 1978 to 2008. At the same time, multi-resources satellite image

英文关键词： wetland simulation；China wetland change；wetland remote sensing；time series；driving factors