中文摘要：

在基于贝叶斯网络的遥感参数反演模型中引入虚拟节点,提出不确定性知识表示与知识更新方法,推导了遥感反演中的参数后验信息更新公式.用北京顺义遥感实验的多尺度数据对冬小麦的叶面积指数和叶绿素面质量进行反演计算,验证了多知识源反演过程中不确定性知识对参数后验分布概率的调整模式.

英文摘要：

A method on uncertainty knowledge representation is proposed by introducing the concept of virtual node into the Bayesian network while retrieving the land surface vegetation parameters. On the issue that the posterior knowledge is updated, the formulation is deduced to reflect the fact that the retrieved information is the result of knowledge accumulation through adjusting the post probability distribution according to the likelihood rate from the knowledge of current stage. The validation work is carried out by retrieving the Leaf Area Index （LAI） and content of chlorophyll a and chlorophyll b （ρA） using the multi-scale remotely sensed data in the area of Shunyi district in Beijing, north of China.