项目名称: 基于机器学习和融合算法的全球陆表植被覆盖度估算方法研究
项目编号: No.41301353
项目类型: 青年科学基金项目
立项/批准年度: 2014
项目学科: 天文学、地球科学
项目作者: 贾坤
作者单位: 北京师范大学
项目金额: 25万元
中文摘要: 植被覆盖度是刻画地表植被覆盖的一个重要参数,在全球变化研究、地表过程模拟和天气预报数值模拟中发挥着重要的作用。本项目在面向全球植被覆盖度遥感估算方法研究的高空间分辨率样本数据集建设的基础上,研究基于机器学习和融合算法的全球长时间序列、高时间分辨率、高精度的植被覆盖度估算方法,为生产高精度全球植被覆盖度产品,进而改进关键陆面过程的参数化方案及同化技术,开展全球变化应用研究提供数据基础与技术支撑。本研究主要内容和方法包括:(1)全球高空间分辨率植被覆盖度样本数据集建设;(2)基于机器学习算法的全球陆表植被覆盖度遥感反演方法研究;(3)多源遥感数据植被覆盖度产品融合算法研究。通过上述研究,解决全球陆表植被覆盖度产品精度不高的问题,提高遥感估算精度,服务于全球变化和陆表过程研究的基础数据需求。
中文关键词: 植被覆盖度;机器学习;融合;全球陆表;反演算法
英文摘要: Fractional vegetation cover is a key parameter for characterizing land surface vegetation coverage, and plays an important role in global change research, earth surface processes simulation and weather forecast models. In this proposal, based on the building of high spatial resolution global fractional vegetation cover sample data set,machine learning and fusion algorithm will be investigated to develop long time series, high temporal resolution and high accuracy global fractional vegetation cover estimation method. This study will provide technic support for producing high accuracy global fractional vegetation cover product, and then improve the parameterization scheme and assimilation techniques for the key land surface processes, and provide data basis for global change research. The main contents and methods of this study include: (1) High spatial resolution global fractional vegetation cover sample data set building. (2) Development of global fractional vegetation cover estimation method based on machine learning algorithm. (3) Fusion algorithm of fractional vegetation cover products from multi-source remote sensing data. This study will solve the problem of lower accuracy of global land surface fractional vegetation cover product and improve the estimation accuracy, and then service to the data basis acqu
英文关键词: Fractional vegetation cover;Machine learning;Fusion;Global land surface;Estimation method