基于高光谱和HJ-1 CCD的水旱地冬小麦叶绿素含量反演

Retrieval of Irrigated and Dry Land Winter Wheat Chlorophyll Content Based on Hyperspectral and HJ-1 CCD Data

叶绿素作为绿色植物光合作用的必要组成成分,其含量的高低可反映作物的长势状况。实时监测植物叶片叶绿素含量的动态变化是监测植物长势的重要环节。以山西省闻喜县冬小麦为研究对象,基于高光谱技术和实测数据,对研究区冬小麦拔节期的叶绿素含量进行定量估算,并在此基础上利用卫星遥感数据对冬小麦的叶绿素含量进行反演,以达到仅应用卫星遥感数据估测叶绿素含量的目的。结果表明,水旱地冬小麦叶绿素含量敏感波段在可见光区域不同,在近红外区域一致;水旱地分别以DVI和NDVI为变量所构建的预测模型效果最佳,R2值均达到0.9以上,均方根误差分别为0.470 0和0.458 7;对叶绿素含量反演值与实测值对比分析,水地反演值与实测值大致吻合,而旱地反演值则偏高;采用均方根误差(RMSE)法,检验反演值和实际值的符合度,水地RMSE为0.926,旱地RMSE为1.540。

Chlorophyll is an important component of green plants for photosynthesis, and its content can reflect crop condition. Live monitoring dynamic changes of plant chlorophyll content is an important link of its growth monitoring. The winter wheat in Wenxi Shanxi as study subject, based on hypempectra land measured data, was estimated quantitatively the chlorophyll content of winter wheat at jointing stage. On the basis, the chlorophyll content of winter wheat was inverted using satellite remote sensing data, aiming for satellite remote sensing data estimating chlorophyll content. The result showed that bands sensitive to chlorophyll content were different in visible region, while they were the same in near-infrared region between irrigation and dry land. The effect of predicting model was the best for DVI and NDVI as independent variable to irrigated and dry land, R2 reached above 0.8, in addition, RMSE were 0.470 0 and 0.458 7. Compared the inversion with measured values of chlorophyll content, it showed that the two values were roughly coincided with each other in irrigation land, while the inversion value was higher than the measured value in dry land. Using RMSE to test whether the two values are in conformity, it displayed RMSE was 0.926 in irrigation land, and 1.540 in dry land.