利用统计和物理模型反演植被叶片生化组分

Retrieving Foliar Biochemical Components with Statistical and Physical Models

通过比较统计模型、物理模型以及两者的联合模型在反演鲜叶片生化组分含量时的效果 ,结果表明 ,对于叶绿素和水份含量 ,物理模型的反演效果较好 ,对于蛋白质、纤维素 +木质素含量 ,统计模型的反演效果相对较好 ,由物理模型改造得到的三种联合模型 ,能在一定程度上提高物理模型反演蛋白质、纤维素 +木质素含量的精度 ,但和利用统计模型反演这两种生化组分的结果比较 ,改进作用并不明显。对于不同样本组叶片生化组分含量 ,不同模型反演效果均存在一定差异 。

Based on comparison of retrieving foliar biochemicalcomponents with statistical,physical and their united models, the results show that physical model is better than statistical one in retrieving chlorophyll and water contents, and statistical model is better than physical model in retrieving protein and cellulose plus lignin contents. Three united models derived from physical model have increased the precision of retrieving protein and cellulose plus lignin content, but can not achieve better results than statistical model. As for different groups of samples are concerned, there are different effects in retrieving foliar biochemical components especially for statistical model.