项目名称: 结合反问题正则化思想及信息容量方法的南海海面风场多源数据的融合方法研究
项目编号: No.41275113
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: 项杰
作者单位: 中国人民解放军理工大学
项目金额: 80万元
中文摘要: 卫星遥感技术的发展提供了海量的大气和海洋环境数据,如何处理和利用这些数据是一个迫切的问题。数据融合是处理和利用观测数据的手段之一,但数据融合本质上属于反问题,因而求解时面临不适定性的困难。本项目拟开展南海海面风场多源数据的融合方法研究,具体就是利用二维变分同化方法,结合反问题中的正则化技巧和信息容量方法,进行如下四部分的工作:(1)两种海面风场数据(QuikSCAT散射计风场和区域数值预报模式输出的风场)的融合;(2)三种海面风场数据(在上面两种风场数据的基础上再增加微波辐射计AMSR-E风场数据)的融合;(3)海面风场数据融合效果的验证,包括利用独立的测站观测数据进行验证和利用信息容量方法进行验证;(4)两种海面风场数据与三种海面风场数据融合结果的比较分析。本项目旨在探讨变分同化方法结合正则化技巧在多源数据融合中的可应用性,并探讨散射计和辐射计风场数据在数据融合中的作用和影响。
中文关键词: QuikSCAT;二维变分同化(2D Var);融合;正则化;精度验证
英文摘要: Satellite remote sensing technology is a powerful approach of obtaining a sea of data of atmospheric and oceanic environment, and it is a challenging task how to process and apply the sea of remote sensing data. Data blending is one of methods by which a large quantity of data can be processed and utilized. In essence, however, data blending problem is one kind of inverse problems, which means that data blending is an ill-posed problem. In the present project, studies on blending of three kinds of surface wind data over South China Sea are to be fulfilled through the 2-D variational assimilation, along with the regularization technique and theory of information content of observations in inverse problems, which includes the following four parts: (1) blending of two kinds of sea surface wind data, QuikSCAT scatterometer observations and regional mesoscale atmospheric model simulation surface wind data;(2) blending of three kinds of sea surface wind data, QuikSCAT scatterometer observations, radiometer AMSR-E observations and regional mesoscale atmospheric model simulation surface wind data;(3)validation of blending of sea surface wind data through two methods, comparison with independent in situ observations and theory of information content of observations;(4)comparisons between blending of two kinds of sea surf
英文关键词: QuikSCAT;two-dimensional data assimilation by variation;blending;regularization;validation